Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum


Message Sort: Post Order or Asylum Reverse Threaded

BDT Preamp Schematics

67.164.79.218

Posted on May 10, 2008 at 14:34:45
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
Here are the schematics for the BDT preamp.

There is one tube per channel, the 6AR8 Beam Deflection Tube. It works using electric field deflection of an electron beam, similar to the CRT in a scope. The beam gets pushed back and forth between two plates, forming a balanced output.

The outputs go through a custom transformer from Dave Slagle designed for this circuit. The output can be wired up in several ways. The COMs can be wired together and connected to ground, or left separate. You can use one output for SE or both for balanced.

If you look at the spec sheet for the tube the plate curves look perfectly horizontal, this is true with respect to the GRID, but NOT for the deflectors. In order to get horizontal plate curves for the deflectors the plates need to be cascoded. This could be done with triodes or pentodes, but I'm lazy and did it with MOSFETs. Previous experiments showed that this should wok very well. It did. The bias for the cascode is provided from a LND150 cascode CCS across resistors. (as in the Gary Pimm SBCCS) The bias voltage can drive many cascodes so you only need one for as many channels as you have.

The accelerators are driven with 250V and the transformer primary (B+) gets 130V. The B+ can be anywhere from 24mA per tube to less than 1mA dependant on the volume control setting.

The HV supplies are simple C filters driving a regulator. The regulator is based on the infamous Gary Pimm SBCCS circuit I like so much. Its exactly the same except for one extra LND150 in the reference circuit which works as an error amplifier. I've tried many different regulator circuit including shunt types and this was the best sounding. I've tried this in several different designs and it has worked very well. It can easily be scaled up to handle high currents so you can even use it for power amps.

There are two supplies, one for 130V and one for 250V. They are the same except for R3, see the schematic for the two values. The exact voltages are not critical, if they are off a little it will still work fine. The pots help take care of the manufacturing tolerances in the MOSFETS.

The heaters take 6.3V DC (it hums too much with AC), I just used a LM340 type regulator for this. I'll leave that as an exercise for the reader. Pin 5 of the tube MUST be grounded. It connects to a number of internal structures in the tube that have to grounded, as well as the heater. (I think they ran out of pins in the 9 pin base)

I used the Antec AN-1T250 toroidal transformer, with the primaries configured for 240V but actually running at 120v. I stacked the HV secondaries, the bottom driving the 130V supply and both driving the 250V.

I built my very first quick and dirty version of this design many years ago with a 9V battery for the volume control and have never changed it since. So this version is STILL run from a 9V battery. I should run it from a supply but I haven't bothered yet. The battery works great, I have to change it every couple months, but that hasn't been a problem yet.

Currently I'm playing around with different circuits for implementing a balance control. Since the volume is DC controlled adding balance does not harm anything. I haven't yet come up with one I like yet so I didn't put it in the schematic.

The BDT is very sensitive to magnetic fields, hence the rather extensive efforts at keeping down fields from the power supply (which is why there are no chokes) and careful shielding. Its not an insurmountable problem, but it has to be taken into account.

One more thing, because its DC volume control, its easy to build a remote volume. A pot on the end of a wire works for a wired remote. I'm working on a remote that has volume on both the front panel and remote. It uses rotary encoders that send pulses which get counted by a FPGA which drive a DAC which generates the control voltages. There is no free running clock, the pulses from the encoders do all the counting etc, when the knobs are not being turned all the signals are DC. I haven't decided if I'm going to display the volume with Nixie tubes or a magic eye tube. Maybe Nixies for the volume and magic eye for the balance?

John S.

 

Hide full thread outline!
    ...
Eagle pcb, posted on May 10, 2008 at 14:54:12
nc
Audiophile

Posts: 324
Location: San Francisco, CA
Joined: November 19, 2003
Having used CadSoft Eagle to draw the schematic, did you layout a PCB also?

 

RE: Eagle pcb, posted on May 10, 2008 at 17:52:46
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
An old version of one of the power supplies was done as a board with Eagle, it doesn't match the current design though, its been modified since then. The other supply I started from a Gary Pimm board and modified it. The heater supply and cascode board were hand wired on perfboard. The tubes were point to point wired on a separate suspended sub-chassis.

I suppose I could do all or part of the design as a board if there is enough interest. I like laying out boards and these are pretty easy. Would people like the tube sockets on the boards or have the boards separate and wire then to the tube sockets?

John S.

 

RE: Eagle pcb, posted on May 10, 2008 at 18:07:56
nc
Audiophile

Posts: 324
Location: San Francisco, CA
Joined: November 19, 2003
I say tube sockets on board. More idiot-proof.

 

RE: BDT Preamp Schematics, posted on May 12, 2008 at 15:36:55
dhsettim
Audiophile

Posts: 169
Joined: January 10, 2002
Thanks John, your postings are always very interesting. Could you please tell a little more about the GP CCS drived PS. I guess it is a shunt reg but some advice on how it works and for setting it up for use in other projects would be appreciated. I have had great success with Gary's CCSes used on my nuvister and 6H30 converted Golden Tube SE40. regards, tim

 

RE: BDT Preamp Schematics, posted on May 13, 2008 at 09:45:53
kenev
Audiophile

Posts: 148
Location: Athens
Joined: September 19, 2001
Hi John,

Just another of your excellent "weird" designs (actually an improved version of your previous BDT preamp). Glad to see the new, revised circuit. I had purchased a bunch of 6AR8 tubes a couple of years ago, when I first saw your BDT preamp, but never sat down to work on it. I think now is the time for it, although this version is a bit more complicated than the previous one.

Just a couple of questions:
What is the bias voltage for the cascode? (I suppose that trimmer R13 is for setting the bias voltage). Also, do the IRF820s of the cascode circuit need any heatsinking?

Regarding the OPT, what is the primary impedance? As I have built a DIY winding machine and I wind my transformers myself, do you think that Dave Slagle could give me some more information regarding the design of this OPT?

Regards,
Evangelos

 

RE: BDT Preamp Schematics, posted on May 13, 2008 at 13:18:39
johs
Audiophile

Posts: 81
Location: Haugesund, Norway
Joined: April 22, 2003
A long awaited schematic John, thank you! Have collected parts and planned for a BDT preamp with a Gary Pimm phono and was about to start building when the good news came. There are some questions I would like to put forward; Is the BDT tube aspesially sensitive to magnetic fields or would any or most tubes benefit from the same treatment you have given the BDT's? As for the aluminum chassis, will it have to be without ventilation? I am trying to understand the CCS/voltage regulator and the CCS for the cascodes. Gary's CCS's all have a current limiting resistor, R1. I can't find this resistor in your circruits, does this mean that there is no or a different current limitation? What is the purpose of R13, to set a maximum current? What is the maximum amplification with the cascodes?

Johannes

 

Voltage regulator details, posted on May 14, 2008 at 13:59:17
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
The voltage regulator circuit shown is a series pass regulator, not a shunt. Its based on the Gary Pimm CCS with the adition of one extra MOSFET. First a little about how the CCS works and how this makes a good starting point for a voltage regulator.

This type of CCS consists of a voltage regulator developing a contant voltage across a fixed resistor, by ohms law a constant voltage across a fixed resistor means there is a constant current through the resistor. So basically everything in the design is already a voltage regulator! As a matter of fact you CAN use the SBCCS as a regulator exactly as is. The voltage is developed across R1 in Gary's schematic. So take out R1 and put the load there. It works.

As a general purpose regulator the CCS has some problems, primarily because its basically a source follower without any feedback, the output impedance is not very low, it doesn't regulate all that well with changing load currents. What I have done is add Q5 which acts as an error amplifier, it compares the output voltage to the reference voltage and drives the source follower such as to maintain the output voltage matching the reference. This radically loweres the output impedance and due to the high bandwidth of Gary's CCS design the impedance stays low over a wide frequency range.

Next I'll go over some details of how the circuit works and why its a fairly neat implementation. The following is refering to my schematic posted above, not to Gary's circuit, the part identifiers are completely different, although the arrangement is the same except for the one new transistor.

On the right side is a cascoded source follower. The voltage on the bottom (source) of Q1 follows the voltage on its gate minus about 4 volts or so. (if the gate is at 100V the source will be at 96V). The exact difference between the gate and the source will vary somewhat depending on the current flowing through it and the voltage across the transistor. The second is handled by Q2. Its called a cascode, it maintains a constant voltage across Q1, any difference in the input voltage is developed across Q2, thus preventing it from changing the gate to source voltage of Q1.

The left side is in charge of producing the voltage references to drive the gates of Q1 and Q2. This is done by Q3 and Q4 which form their own high quality low current (0.33mA) CCS. This current flows through R2, R3 and R8 developing constant voltages across them. (R1 is part of the CCS) The voltages can be changed by changing the resistor values.

R3 and R8 comprise the main voltage reference. They are filtered by C1 to make a low noise reference. R2 is the cascode bias. It determines the voltage difference between Q1 and Q2. As described so far this is the original Gary design. The Q3/Q4 CCS performs two functions, main voltage reference and cascode bias.

The modification to a feedback regulator is done by adding Q5. This is basically a comparitor, one input is the reference voltage underneath it, the other is the output voltage. The "output" is the drain of the MOSFET which drives the gate of Q1. The beauty here is that the same CCS that provides the consant current for the references is also the CCS load for the error amplifier! The same CCS is being used for three different tasks. Most error amplifiers just use a resistor load for the amp, but they work even better with a CCS load, but that is rearely done because of the extra complexity of the CCS load. Because one was already in Gary's design it was a piece of cake to have a CCS loaded error amp.

This error amp takes care of the changes in gate to source voltage caused by current fluctuations from the load. The result is very good regulation no matter what the load is doing.

Why didn't Gary do this in the original CCS design? Because its not really needed in a CCS. The problem is caused by changing current through Q1, but in a CCS the current through Q1 is constant so it doesn't matter. When using it as a general purpose regulator it DOES matter, so I added the error amp.

If anyone wants to use this as a voltage regulator for other purposes make sure you include C6 and R7. Without them the regulator will oscilate up in the MHz region somewhere. Especially if you convert an existing board of Gary's make sure you add C6 (R7 goes in place of Gary's R1).

Well thats it, I hope thats not TOO much detail on how this works. I've tried this in many different places and its the best sounding regulator I've ever tried, including a bunch of different shunt regulator designs.

John S.

 

RE: BDT Preamp Schematics, posted on May 14, 2008 at 14:59:43
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
Hi Johanes,
the BDT tubes are much more susceptible to magnetic fields than most tubes, so I had to go out of the way to lower those fields. With the metglas shield in place I don't know if the other techniques would be necessary or not. I haven't tried it with the original PS or without the heavy aluminum box and just the shield.

I'm running it without ventilation holes but thats not necessary. The more holes the more you risk fields getting in which is why I didn't ventilate it in the first place. The whole box acts as a heatsink so the air inside really doesn't get all that hot.

See the post on how the regulators work, I hope that will answer the questions. R13 is just a way to vary the cascode bias voltage. The manufacturing tolerances on the threshold of the LND150s are not very tight so the current from the CCS they form can vary a bit from nominal, thus the variable resistor to "tweak" the cascode voltage.

The amplification depends almost entirely on the load impedance. The maximum transconductance is about 1mA/V (1000umhos or 1mS), so with a load of 100K you can get a gain of 100 (40db). You could go even higher if you are feeding into a very high input impedance. I keep on having a death wish and want to try this as a phono preamp!

Thanks for the interest,

John S.

 

RE: BDT Preamp Schematics, posted on May 14, 2008 at 15:12:35
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
Hi Evangelos!

I use a Cascode bias of 100V with a B+ of around 130V or so. The exact value is not important.

The maximum idle current through a plate is about 12mA, so at 30V drop across the cascode MOSFETS they are only disipating 360mW, which is fine without heatsinks.

On the transformer front all that will have to get answered by Dave, I really don't know the details.

As to the complexity, most is in the PS. You can certainly try building it with a different PS configuration. I tried a bunch and this sounded the best, but that doesn't mean something else is going to sound BAD.

My original version had both accelerator and plates running at around 180V, this was really too high for the plates, they got too hot at maximum current. With 100V on the plates they stay much cooler. The tube was designed for an accelerator voltage of 250V and it really does work best there. You can certainly run it lower, but then the maximum current is less. This is important to getting a wide gain range. At very low currents the tube will start to distort on high level inputs. By keeping the max current (ie max gain) high you can still have the minimum gain be quite low before the distortion sets in.

Thanks for the interest,

John S.

 

RE: BDT Preamp Schematics, posted on May 15, 2008 at 02:41:48
kenev
Audiophile

Posts: 148
Location: Athens
Joined: September 19, 2001
>>Thanks for the interest<<

Hi John,

In fact, I'm the one who should thank you for your brilliant and innovative designs! And, also, thank you for spending your time to explain things (for the least electronics experts, like me). For two years now, I'm happily listening through a KT88 transconductance amp. This is the best sounding amp I have tried with my Fostexes. I shall give this preamp a try soon (I hope).

Regards,
Evangelos

 

RE: Eagle pcb, posted on May 15, 2008 at 16:59:59
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
Thanks John,
Like a lot of posters here, I bought some 5AR8 tubes a couple of years ago meaning to give it a go but never got around to it - now's the time. I'm planning on building a Zeus amp (Susan Parker's design) & I was agonising over the best way to feed a balanced signal into a 600ohm input trafo but now this may well allow me to dipense with the input trafo & use this BDT preamp as both vol control & phase splitter.

Waht do you think - is this feasible? How would I bias the Mosfets if I substituted the BDT preamp in place of the I/P trafo. Sorry for the questions but as you can see I'm no Ee & struggle with the simplest of things!

 

RE: Eagle pcb, posted on May 15, 2008 at 17:02:36
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001



Forgot to include image & I would love a board with sockets on board!

 

RE: BDT Preamp Schematics, posted on May 15, 2008 at 17:13:32
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
What price is Dave Slagl's trafo for this preamp? I was going to ask what else comes near in sound quality but I seem to remember you saying the trafo elevated the sound quality a lot. Anything else come close?

I suspect price + shipping to Ireland is going to be more than I can afford!

 

Thanks John, posted on May 15, 2008 at 17:27:36
dhsettim
Audiophile

Posts: 169
Joined: January 10, 2002
Much appreciated, tim

 

RE: BDT Preamp Schematics, posted on May 16, 2008 at 08:13:20
jkeny


 
Can IRF1820 be substituted? I can't find it anywhere or it's datasheet! What would be a suitable substitute & what component value changes would it entail?

 

RE: Voltage regulator details, posted on May 16, 2008 at 13:34:16
cv


 
Hi John,
Very elegant circuit! And as elegant an explanation, though I was wondering about how you determined the value of C2. I've tried simulating rejection vs frequency and noticed there is some play between the various time constants in there. To be honest, I'm getting some slightly odd results and need to double check my sims and mosfet models.

Did you try the circuit without C2 at all? Wouldn't be a cascode at ac without it I suppose, as the load is on the source of Q1 rather than the the drain of Q2.

Also wondering which shunt regs you'd compared your circuit to; I've contemplated Emile Sprenger's and (similar) VSE the Camille designs, but yours would be far less hassle, not to mention efficient. Thanks very much indeed for sharing this one!

Best,
Chris

 

RE: Voltage regulator details, posted on May 16, 2008 at 15:54:30
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
I determined the value of C2 by trial and error! Gary originally had a 100uf cap there. I think the primary reason for the large value was to provide a long time constant for a "soft start" effect. I found that when using it as a regulator it would oscilate like mad during that startup phase. You could hear it in the amp as a tone that would start at a low frequency and work its way up through the audio spectrum and eventually go away. At this value the time constant is small enough that it doesn't go crazy during startup. I left it there!

At very small values (or not there at all) the cascode bias has a high impedance to the output, which causes it to oscilate as well. The 2.2uf seems to keep everything nice and happy.

My shunt regulators I tried were not specifically any one persons designs but a number of variations I came up with myself using both pentodes, bipolars and mosfets and various hybrids of the types. Trying different CCS designs and different shunt circuits. This circuit beat them all sonically and is quite a bit simpler than some of the shunt regs.

John S.

 

RE: BDT Preamp Schematics, posted on May 16, 2008 at 16:01:08
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
Many different MOSFETS will work fine for Q1 and Q2. Do not substitute for the LND150s. You want enhancement mode N channel at a voltage rating that can take anything you will be throwing at it. For most tube circuits that are running under an amp of current you want to use a power mosfet designed for relatively low current, ie in the 1 amp range not the 40 amp range.

I personally have used at least six different part numbers and they all have worked fine.

John S.

 

RE: Eagle pcb, posted on May 16, 2008 at 16:10:52
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
It would work perfectly in this situation. The ground connection in the BDT schematic gets set to the bias voltage in your amp schematic, definately no need for the extra input transformer. You can adjust the load resistors after the transformer in the BDT schematic to be whatever value you need to get the gain to drive the amp. The BDT circuit can generate high levels of gain if you need it.

Using it this way makes a wonderful integrated amp, I'm actually going to be doing something similar (but with pentodes).

John S.

 

RE: BDT Preamp Schematics, posted on May 16, 2008 at 17:49:03
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
Aha, I think the schematic shows IRF1820 rather than IRF820 - this is what confused me!

 

RE: BDT Preamp Schematics, posted on May 16, 2008 at 18:25:53
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
I emailed Dave Slagle for prices & any spec info he was willing to provide - just got price of $350 a pair & it's wound on a nickel core.

I can't afford them at this price so I'm wondering what could substitue for this - Im thinking that a PS toroid might work here - I've heard these used as input & output transformers with great effect.

Any ideas - what impedance does the SAR8 tube like to run into?

 

RE: BDT Preamp Schematics, posted on May 16, 2008 at 19:41:50
C.Y
Audiophile

Posts: 342
Location: S.Korea
Joined: September 7, 2002
Thank you for sharing this!
I am really intrigued of the possible sound but I wonder should I use Slagle transformer or Sowter 9050 that I have.

BTW, I know that you are using parafeed transconductance SE amp for your Lowther. Recently I tried little Amveco PCB power transformer for parafeed and I really liked it. I tried many other parafeed OPTs and this cheap Amveco toroid sound no less than them.

Cheers,
C.Y

 

RE: BDT Preamp Schematics, posted on May 17, 2008 at 00:24:04
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
Aha, the "I" stands for "Insulated", the tab is insulated from the pins so you can screw it into a heatsink without needing mica insulators, nylon washers etc. Especially with the high voltages floating around these circuits I like to not have the HV where I could touch it!

John S.

 

RE: BDT Preamp Schematics, posted on May 17, 2008 at 00:41:47
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
Because of the cascoded topology it has a very high impedance (megaohms), the gain is determined by the load impedance. You figure out what gain you need, assume a transconductance of 1mA/V, that determines the load impedance you need on the transformer primary. Use a transformer and secondary load value that give presents the calculated load impedance to the circuit.

Example, a load of 10K gives a gain of 10X (on each output). If you have a 1:1 transformer you want a 10K load on the secondary, If you have a 3:1 transformer you want about 3.3K on the secondary (impedance is turns ratio squared, hence 9X the impedance) (this gives 30K load (hence 3 times higher gain) then the divide by 3 in the transformer gives the same overall gain)

You can play around with all kinds of different transformers and impedances and see what sounds the best to you.

Just be careful, many transformers are not designed to have high impedances on both the secondary and primary, thats one of the reasons for the custom transformer, it works best with high impedances.

I've used the cheap Hammond 124B, Lundahl LL1660 and Dave's. Dave's sounded significantly better than the other two, but you can still get quite good sound out of this even with the 124B, just not as good as its ultimately capable of.

John S.

 

RE: BDT Preamp Schematics, posted on May 17, 2008 at 00:57:11
John Swenson
Audiophile

Posts: 2422
Location: No. California
Joined: October 13, 2002
You can certainly try this with the Sowter, nothing stopping you from doing so. Just carefully work out the load impedance based on the gain you need and the ratio.

Since the circuits impedance is very high there is no specific impedance you have to aim for, use whatever gives you the gain you want given the transconductance of 1mA/V.

Say you want a gain of 5, that means a load of 5K if it was a 1:1, but with the 6:1 you want it 6 times that or 30K. Thus the load impedance is 30K/36 = 833 ohms, pretty close to what the transformer was designed for. If you want more gain increase the load resistor, for less decrease the load resistor. Just make SURE you put an explicit load resistor across the secondary, that is what determines the impedance the circuit sees. If you don't put one in there the load will be the 50K impedance of the power amp which will be X 36 from the transformer which is 1.8 megaohm, which is a gain of 1800 for the tube / 6 = 300 for the whole shebang. I think that would kind of overload things! (and probably oscillate all over the place)

The only issue with a 6:1 is to make sure you have enough voltage for the PP swing you are going to need. With the voltages as speced that comes out to about 100V PP maximum across th primary. With a 30X gain thats a maximum input signal of 3.3V PP for full output. That might be fine, just be aware of it.

John S.

 

RE: BDT Preamp Schematics, posted on May 17, 2008 at 04:18:51
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
I'm sorry John, I didn't spot the "I" - interpreted it as a "1", doh!!!

 

RE: Voltage regulator details, posted on May 17, 2008 at 04:31:12
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
Thanks John for all the details & guidance you have provided - I think as you can see from the volume of replies that there is a big interest in this unique pre-amp.

I'm going to try this circuit integrated into a "Baby Huey" power amp & later a Zeus amp. Two quite dissimilar but also unqique designs.

I'm wondereing how this Voltage Regulator can be scaled up for a power amp as I thought of using it for both duties - preamp & poweramp VR?

 

RE: Voltage regulator details, posted on May 17, 2008 at 15:31:41
cv


 
Good old fashioned empirical approach then!
Thanks John - you're s a scholar and a gentlemen
cheers
cv

 

RE: Power Amp Voltage regulator details, posted on May 17, 2008 at 16:19:36
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
I know there are high voltage/high power Mosfets available such as 350V/12A FDPF12N35 for reasonable prices (digikey $1) but how would LND150 be made suitable?

 

RE: Eagle pcb, posted on May 17, 2008 at 19:15:56
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001



OK, just to make sure I got it - The output trafo of the BDT preamp replaces the input trafo of the Zeus amp - the Com+ & Com- of the BD preamp are tied together & get the bias voltage (~4.5V) applied here to bias the Mosfets of the Zeus amp?

I also asked a question above about how the VR might be scaled up for power amp use - I'm thinking of integrating this preamp onto a "baby Huey" circuit I have and having a preamp out option for feeding my Zeus.

The feed from the BDT to baby huey will probably be SE and not as elegant as the differential input into the Zeus

 

RE: Eagle pcb, posted on May 18, 2008 at 04:46:04
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
One final thing, could this VR be made into a High current/low voltage regulator - I need something along the lines of 8A/40V - burst not continuous & I need to be able to vary the voltage for a voltage controlled volume adjustment?

 

RE: BDT Preamp Schematics, posted on May 19, 2008 at 00:47:20
C.Y
Audiophile

Posts: 342
Location: S.Korea
Joined: September 7, 2002
Thank you for the kind post.
Four 6AR8s arrived in today and its time to warm up the work bench...

Cheers,
C.Y

 

RE: BDT Preamp Schematics, posted on June 1, 2008 at 20:02:55
Gary P
Audiophile

Posts: 754
Joined: March 30, 2001
Being intrigued by the regulator circuit John came up with I built one up and proceeded to test, measure, and blow it up many times in the interest of learning more about the circuit.

Performance is very good the way John has it set up but there are improvements that can be made both for performance and reliability.

I changed a couple of capacitor values, one for improved stability, the other for better performance. I also added some protection circuitry and a current limit circuit.

I'm planning on building several of these into a multiple output bench supply to replace my old Tektronix Engineering power supply. The Tek supply works pretty well but is noisy, both electrically and acoustically. It also has a tendency to blow up at inopportune times as it does not have good current limiting, instead, relying on 2 watt CC resistors on the SS rectifier diodes. It let loose again today during testing of the Swenson regulator. Had to open all the windows to get that wonderful burnt CC smell out of the house. Maggie was not very impressed with the smell but enjoyed the show... It was interesting to see the meters on the power supply back lit by orange flame.

I saw John's preamp at VSAC but never found John amongst the crowds.

Anyway, take a look at me website for the info.

 

Fascinating!, posted on June 3, 2008 at 06:29:30
JoshK
Audiophile

Posts: 790
Location: NJ/NYC
Joined: August 3, 2001
You and John amaze me with the things you come up with. I really want to try out this regulator now.

 

RE: BDT Preamp Schematics, posted on June 3, 2008 at 13:12:35
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
I agree, JS & GP are both wonderfully productive, creative & technically astute people. I'm in the final stages of building this BDT preamp & will be incorporating GP's mods to VR - looking forward to listening soon & will report back!

Has anybody built this one yet & care to share?

 

RE: BDT Preamp Schematics, posted on June 12, 2008 at 07:31:23
kenev
Audiophile

Posts: 148
Location: Athens
Joined: September 19, 2001
Hi Gary,

I read your excellent article with improvements on John Swenson's regulator and I would like to build it and give it a try. Just a couple of questions:

1. Did you use a new pcb or just modified the existing SBCCS pcb?

2. What kind of capacitor did you use for C2? Is an electrolytic OK there?

3. Also, what type of capacitors do you recommend for C1 and C5?

Regards,
Evangelos

 

RE: BDT Preamp Schematics, posted on June 12, 2008 at 18:34:35
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
Evangelos,
1. I did mine point to point with Gary's mods but left out the over-current section
2. C2 is a electro 2.2uF/50V
3. C1 & C6 are 10uF & 2.2uF metallised polyester >250V, I think?

 

RE: BDT Preamp Schematics, posted on June 12, 2008 at 21:14:55
kenev
Audiophile

Posts: 148
Location: Athens
Joined: September 19, 2001
Hi jkeny,

I think we have confused it a little: I was referring to Gary's schematic and not to John's one. I see that the numbering you ae referring to correspond to John's schematic. I also noticed that I made a mistake in my previous post: there is no C6 in Gary's schematic, I had to write C5.

Anyway, I think that a 10uF/250V metallized polyester capacitor should be quite bulky. Could I use an alectro instead?

Regards,
Evangelos

 

RE: BDT Preamp Schematics, posted on June 13, 2008 at 01:10:43
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
The two >250V caps are being used as noise suppressors and carry the full output voltage accross them. I think they need to be film based for their high frequency characteristics but experts may well correct this view. Yes they are bulky!

The electro, on the other hand, doesn't carry the full ouput voltage (I measured about 100V) so 100V electro should do!

 

RE: BDT Preamp Schematics, posted on June 13, 2008 at 16:02:43
mjk
Manufacturer

Posts: 841
Location: Eureka, CA
Joined: October 24, 2007



Hi John,

Great to meet you at VSAC; I'm working on the BDT-based compressor and limiter. I have a couple of questions:

"The outputs go through a custom transformer from Dave Slagle designed for this circuit."

What is the primary inductance of this transformer? I was worried about being able to drive finite inductance at low frequency and at high attenuation when the plate current is ~1mA or less. I guess it's relative to the load resistor used to set the stage gain.

"If you look at the spec sheet for the tube the plate curves look perfectly horizontal, this is true with respect to the GRID, but NOT for the deflectors. In order to get horizontal plate curves for the deflectors the plates need to be cascoded."

I think the question is how linear will the transfer function be? The attached curves show my planned operating range (different tube but same basic circuit). The curves are measured with a fixed 250V anode voltage.

Looking at the pentode curve (not attached but they all look alike), one might expect small impact of anode voltage on current *OR ON CURRENT STEERING LINEARITY*.

However, the BDT/MOSFET cascode is IMHO even cooler than the regulator design and I will certainly give it a try with and without.

Thanks!

Michael

 

RE: BDT Preamp Schematics, posted on June 13, 2008 at 18:19:36
Gary P
Audiophile

Posts: 754
Joined: March 30, 2001
Hi Evangelos,

The test work was done on a modified self bias CCS board. I have a board layout in the works that is mostly done. Only need to order a set of boards so I can stuff one of them and make sure it all works as planned. When all checks out good I'll put the ExpressPCB board file and a BOM in the down load section of my web page.

I've added more to the Swenson regulator page this evening with more info on setting up the regulator, the new board layout, and a picture of the test mule for entertainment.

Using the circuit designators on my schematic, C1 and C2 are film caps, C5 is a 35V electolytic. C1 and C2 need to be high performance caps, C5 is non critical.

 

Thank You, posted on June 13, 2008 at 19:49:14
nc
Audiophile

Posts: 324
Location: San Francisco, CA
Joined: November 19, 2003

 

RE: BDT Preamp Schematics, posted on June 14, 2008 at 23:12:40
jkeny
Manufacturer

Posts: 314
Joined: May 4, 2001
Hi Gary,
Yes again, thank you for this inital CCS circuit & your mods to John Swenson's VR variation.

I notice a change to the value of R14 from 27K in JS's schematic to 33K in your schematic but I see no mention of it in the write-up of your mods.

Is this value significant or is it just what was on hand - I think it's purpose is to biase the gates of Q1 & Q2 with at least 20V difference? I presume with 33K the biase differential between these gates will be greater than 20V?

 

RE: BDT Preamp Schematics, posted on June 17, 2008 at 01:10:12
kenev
Audiophile

Posts: 148
Location: Athens
Joined: September 19, 2001
Thanks for your help.
Gary also added quite helpful information in his relative page.
Regards,
Evangelos

 

RE: BDT Preamp Schematics, posted on June 18, 2008 at 07:45:42
kenev
Audiophile

Posts: 148
Location: Athens
Joined: September 19, 2001
John,

On the transformer issue, I have an important question: does the primary of the transformer carry any DC current or it behaves like a PP output transformer with the DC currents in opposite directions and cancelling each other? In other words, do I have to calculate the Tx with an air gap or with interleaved laminations?

Regards,
Evangelos

 

Page processed in 0.023 seconds.