Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
I've gotten various amounts of conflicting information regarding filter cap values with the 5r4gy and am seeking advice.
I have a pair of amps that currently uses a 5y3 as its rectifier,first input cap is 47uf. I had one die last night and while digging through a box of tubes (as my entire backup supply I usually use is missing) I was happy to find 11 spare 5y3 tubes.
But I also came across some 5r4gy tubes in a ST envelope. They just looked obscenely large and kind of screamed vintage. I decided I had to try and run them if possible.
All the specs look ok, filament is the same as a 5y3...the biggest difference is voltage drop...but I'd be willing to give them a chance anyway.
The problem is datasheets list 4uf as..at least the recommended cap value. Some have said thats a 1948 value and that later books listed this as 20uf. Many have said don't go over 4uf...just as many seem to have said they use a 5r4gy with 47uf (or more) with no problems.
I've gathered one of three things will happen if I slap these in: they'll just work, they will be fine at first but prematurely fail, they'll start arcing internally. Of those three...I think the arcing is the most fatal but only to the tube.
Thanks in advance.
There were too many responses for me to respond to personally. But I would like ti give everyone a huge thanks. I am not an engineer and don't pretend to be one, though I've picked some of it up from my uears of slinging solder in jukebox amps.
I now know specifics of tube power supplies more than I did 48 hours ago...by no means an expert and far from it...but at least I now have the right path.
I put the rectifiers in my amps last night and there were no ill effects. Ran them for 5 hours and they sounded great. Look even better having those huge hanging filament rectifiers there.
Yeah...I got silly and wanted to run the things because of how they looked.
And yes....I checked the tubes in my Precision 612 before attempting to use them.
I saw in your system description that you're using a pair of Motorola HS-619 mono's, so I'll assume that these are the amps in question. I did a quick search and found a schematic online that shows the amp to have a current draw of 90 mA and the voltage at the input of the filter as being 345V.
Based on these figures I ran a PSUD sim which shows that the peak current per cycle (peak repetitive) to be 350 mA using a 47uF cap at the input to the filter. This is well within the 5R4GY's safe operation envelope, since both current and B+ voltage are on the lowish side for this amp. As I had indicated in my previous post the max steady-state peak current for this tube is 650mA.
As a comparison, changing the first cap to 4uF reduces the peak current to 311mA - not that big of a drop, but also reduces the B+ at the output of the filter to 332V vs. 344V in the case of the 47uF cap, all else being equal. Hope this helps.
I attempted to write TWO replies last night and neither one showed up.
After your quick explanation and screen shots, in addition to hammering my brain to recall old information....things started to make a little more sense.
Stuck my nice 5R4GY tubes in the amps and ran them for several hours. Improvement? Not really...I just wanted some cool looking rectifiers.
Thanks again for the extra effort of mocking that up in PSUD. That looks like something I'll need when I start attempting to design/ripoff a quad 6l6 amp.
Glad it worked out. In retrospect, I should have taken a look at the 5Y3 datasheet to see that it has a 440mA peak repetitive current rating which is two thirds that of the 5R4GY. That would have put any concerns to rest since the 5Y3 has been running in your amp for a long time without issue.
At the very least, now you know how to check for compatibility with other rectifiers in this or other amps. PSUD might have its limitations, but is a very nice and simple utility for stuff like this.
In the future, if you ever get the inclination to do a mod/upgrade, you might want to consider replacing the 120 Ohm series resistor with a filter choke such as the Triad C-17X. It's a small choke and fairly inexpensive and I think that it would offer a decent improvement in sound over the resistor.
The schematic I found at the link below shows the PT secondary winding at 15 ohms per side of the center tap.
With 47uf the peak current increases to 370mA. Still well within the tubes rating of 650mA.
However if you kick the voltage of the power transformer up to 900vac center tapped and increase the current draw to 200mA. the rectifier peak current becomes 759mA.
The input cap needs to decreased to 4uf to keep the peak current below the rated 650mA level.
The max uf value for the input cap depends on the 'voltage, the current draw and the PT secondary DCR' of the particular circuit one is using the rectifier tube with.
Have Fun and Enjoy the Music
"Still Working the Problem"
The Sams schematic shows the secondary ad 52ohms and 57ohms each side of the ct of the secondary.
The Beitmanns or whatever they were linking to over at AK I found to be slightly less accurate in some cases than the sams.
"The max uf value for the input cap depends on the 'voltage, the current draw and the PT secondary DCR' "
That's mostly correct; the last part of that sentence should read: "...and the PT source resistance". As you know, that includes the PT's primary DCR and the formula is Rsource=Rsec+N^2*Rpri+Ra. Admittedly, I got lazy and didn't include the primary's contribution in the sim I posted. Had I done that, the PT's resistance would've come out to 74 Ohms. Which brings me to another point; the link you provided (which I can't view for the time being), that has the secondary resistance at 15 Ohms, seems low to me. I have 10 lb trannies that have higher resistance values than this seemingly small-ish PT. Unless someone took actual measurements of the transformer, I would remain sceptical. Bellow, I am including a link to the schematic that I found, showing current, voltage and resistance values for the psu/power transformer.
The other point you brought up about increased current and voltage, I will also partly agree with. Two reasons: one, is that this is a class AB amp and when at full swing one of the two output tubes is in cut-off thus drawing no current. Of course the tube that remains on and depending on the steepness of the load-line could reach three even four times its quiescent current draw; but it's one tube pulling current and even if it sees a five-fold increase in conduction, it should not exceed 200 mA in this amp - we're talking about a 6V6 tube here.
Second, the tube that's drawing max current, will have its plate voltage pulled down to a minimum value. This is to say that the two conditions of max current and max voltage do not occur simultaneously, so kicking up both at the same time is not warranted. When I ran the original sim with a step load, 210mA draw was the limit before peak current at the rectifier reached its 650mA limit. If the schematic I used is the correct one, then you could probably go a bit higher on the current. If the schematic that you used is the correct one, you'd go a bit lower say around 180mA max draw.
In any case, I still think that the OP could probably try the 5R4GY with no ill effects, assuming of course the the tube is in good condition with no gas, shorts etc.
It occurred to me after I posted my response to your post, that you were talking in general terms about possible conditions where a small first cap is necessitated when the 5R4GY is used and not necessarily about this specific application. In that case, I will agree with all you wrote.
I just posted a long response to you post above but after reading this one I deleted it.
Yes I was speaking in general terms, trying to make the point that the number given on the datasheet (any rectifier tube, any datasheet) is dependent on the circuit condition with which the rectified tube is being used and is not a max or "hard" number.
Making sure the peak repetitive current rating (for whatever rectifier tube we are using) is not exceeded is what guides us to the max uf value of the input cap based on the circuit conditions, case by case.
Thanks for being able to "read me".
Have Fun and Enjoy the Music
"Still Working the Problem"
sideliner has this right.
The figure given on the datasheet is not a max or suggested value.
It's a value given the condistion the rectifier tube is under.
That is to say, you can't exceed the peak repeative current rating of the tube.
To know whether or not you are doing that you need to know the voltage, current, value of 1st cap, DCR of all parts, etc.
Then you need to do the math.
The datasheet gives the series resistance per plate (that would be the DCR of the secondary) for the condition under which it states the max uf value.
Again, there is not a "one size fits all" max uf value but there is a "one size fits all" max peak repeative current rating that can not be exceeded without causing damage.
If the DCR of the secondary is higher then you can use larger a cap than the datasheet states. If the DCR is lower then you will have to use smaller cap than the datasheet states.
It's all about the peak repeative current rating of the tube.
BTW a inrush current limiter will solve turn on surge problems but will do nothing to solve the problem of exceeding the peak repeative current rating of the tube.
The peak repeative current is the current that happens every half cycle as the diodes turn on and off.
The larger the 1st cap (and/or the lower the series resistance per plate) the less time each diode will be on, so each diode will have to deliver more current while they are on (to make up for their shorter duty cycle).
Have Fun and Enjoy the Music
"Still Working the Problem"
I use a F&T 10uf/500v cap as the initial capacitor in all my DIY builds.
That way I can try different rectifier tubes, to change my B+, without worrying about first capacitor match ups.
A capacitor input filter ( first element after the rectifier tube ) is the worst possible design to have. The bigger the cap in uF value, the worst - off is the design.
The best practice is to always have a choke, of some kind, with adequate current rating, and lowest possible DCR and mass, as the first element the tube rectifier sees.
Chokes in tube amps, as for example, a 2A3 amp, need to all be 20 Ohms or less in DCR, and there is an advantage to getting down to ten Ohms or less. A double L/C section B+ filter is tops, as in L1/C1/L2/C2 to the Finals.
If employing high drop rectifiers, like a 5R4 or a whimpy 5Y3, you should avoid using larger uF C1 input cap values, its NG, design - wise, on a long-term reliability basis.
That 47 muF. is too big for all the common 5 V. rectifiers, except the 5AR4/GZ34. Check the data sheets, for yourself.
The forward drop percentage in a 5Y3 is greater than that in a 5R4. Remember, the 67 V. drop in a 5R4 is at its 250 mA. rated max. A 5Y3 is rated for 125 mA. max.
A CL-130 inrush current limiter between the vacuum rectifier and the comparatively large 1st filter cap. might save the day.
The 5R4GY has a relatively low steady-state peak current rating of 650mA. That's the reason for the low value specification on the PSU filter's first cap. At relatively low B+ and current draw conditions, the first cap could be considerably higher such as the 47uF value that you have seen. You will not know for certain unless you have the specifics of the power supply circuit and component values. If these are available (or if you could take some measurements) a simple PSUDII simulation will let you know what you can use in the first capacitor position.
Post a Followup:
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: