Possible power supply problems?

My ST-70 clone w/ VTA driver board has developed a flukey fault. A set of EL-34 tubes that would once bias will not make a voltage drop of .4 VDC across the 10 ohm cathode resistors any more. It now makes only .3 volts at the most. The amp has the DynaClone PA-060-S power xformer. (without bottom bell cover) It's a little more than five years old.

That set of tubes will bias in another amp just fine and they also check out OK in my tube tester. (B&K 747B)

I have a set of Valve Art KT-88's that will bias up in the flukey amp.

I thought at first that my GZ-34 rectifier was getting weak because there is only 325 VDC at rectifier pin 8 to ground with the amp up and running. Should be about 435 volts?. I ordered up a GZ-68 copper cap but the B+ remains at 325 V with either.

Another weird thing going on here is with the power supply choke on the under side of the chassis. When I install the bottom part of the chassis the choke is rattling the hell out of it. It seems to be acting like an electromagnet. Remove the bottom portion of the chassis and the buzzes and rattles disappear. Could it be that in the B+ circuit that something is drawing a lot of current thus making the choke behave that way?

B+ winding of transformer measures 762 VAC (!!) without tubes installed. Pin 8 of rectifier to ground is 505 VDC w/o tubes installed.

With all tubes installed, B+ xformer out measures 750 VAC and pin 8 is at 325 VDC to ground. My line voltage here is 117 volts.

I measured the drop across the 2.2k resistor between the last two sections of the can cap at 28.8 V. I believe that would mean that the driver board is drawing .013 amps.

I also checked out the bias supply circuit. Pulled all the tubes and fired it up. Bias tap gives 57.7 VAC and rectifies out to 77 VDC. Checked at pin 6 of each output tube socket while turning the bias pots and measured -28 to -54 VDC. Checked out all the 10 ohm cathode resistors too.

Something is amiss here and I'm having trouble putting my finger on it. Could it be that the PA-060 is getting tired? Output transformer problems? Don't know what to check next. Any and all help and instruction will be followed completely, carefully and will be greatly appreciated.

Chip

Hi Bob:

Thanks for replying to my post. I'm gonna check out that can cap. I'll post back so all will know the findings and outcome. I don't have a spare cap on hand so if I have to order one it may be a while.

Thanks again and nice to meet you.

Chip

You could do what I did... Just say "no" to quad caps As a debug tool I went to the local guitar amp repair place and bought a couple 47µF, 350V caps to bypass the "first" section of my suspect quad cap. That stopped the distortion and kept the 3A slo-blo from opening. Then I knew what needed to be done. I measure 525V steady state on pin #8 of my 5AR4 so a 525WVDC quad cap seems like asking for trouble in my case. I built the cap board below (actually two--one for each of my Mark IIIs) and have never looked back.

Do you have a capacitance meter? FYI: I have found the $40 DVM from Harbor Freight with built in 0-200µF cap meter to be a very useful tool. Not quite sure what value the relative humidity and ambient temperature displays add... Fry's sells the same meter (relabeled) for $50.

OK... All fixed up now. Making beautiful music again better than ever. Indeed I did just say "NO" to multi-section "canned caps". I bought a new and improved (Re-sized and no more surface mount components)blank cap board from Triode Electronics and a bunch of caps and resistors from Mouser. This cap board still has a bias supply section that I just skipped since I have the VTA driver board. The reason I picked my own parts for the board was because the Triode E board parts only rate the B+ section to 500V AND it would have came with bias supply components which I didn't need. Two 250WVDC caps in series, maybe a little low. I upped the ante on the WV ratings and the capacitance per section is now 41-75-75-75 @ 900V for the first three sections, 700V for the last section. Triode board with parts would have been 40-195-135-135 @ 500V. I had to stay within physical diameter and length dimension limits for the individual electrolytic components and that's one reason it wound up like it did. I was just afraid that 500V rating would not provide enough headroom in the voltage handling department. Headroom in spades now. Should last a good long while. The only down side I see to going the way of the board instead of the can is COST.

Chances are the can cap that was in this amp was bad from the go since this amp had always sounded a bit thin and got thinner with time. I'm glad it finally gave up what it had left of a ghost and made me do something else. This '70 ain't EVER sounded this good. My '35 always blew it away handily. No more though. It's like I'm listening to it for the first time and I don't want to turn it off.

Link below to pics.

http://cid-6f6493fd34a81b0b.skydrive.live.com/redir.aspx?page=play&resid=6F6493FD34A81B0B!103

Excellent! I looked at the pictures, that's a nice assembly job. I built my two boards for about $18 each. They measures 25-50-25-25uF, using series 47uF, 450WVDC caps and 220K balancing resistors. I am no fan of the quad cap "spam in a can" With the higher voltage rating I now have I am thinking about silicon rectification and KT-88s in the Mark IIIs.
...j