First build B+ is low

[Luthierwnc]

Hi Guys;

I am wrapping my first amp in this style and have run across an issue I haven't seen in other amps. I am using a Heyboer Marshall JTM PT that has 700 VCT. The schematic is along the lines of the Skyliner redux (HRM S3988009.2) but with a boost. No jazz or deep switches. Under load it can only manage 412 VDC on the plates. I have a pair of Phillips 7581s biased at a very modest 34 ma per side (shunt measured). I was expecting more like 440 VDC. The wave patterns are very clean, it runs cool and is quiet as a mouse. AX's in the first two holes and an AT in the phase inverter.

I haven't ever built an amp with this type of heavily regulated power supply and am wondering if I am sending too much juice to ground across all those series caps. The supply is identical to the Skyliners except for pulling the bias off one of the HV taps. Here are all the supply voltages and feed resistor values:

V1B 206 150k

B+5 327

V1A 208 220k

V2B 209 120k

B+4 343

V2A 225 180k

V3B 273 110k (+- 10k trimmer)

B+3 388

V3A 272 100k (+- 10k trimmer)

Screens 411

Plates 412

The amp sounds really great for blues and rock but for more subtle playing I could use more headroom. Any power supply comments would be appreciated.

Thanks, Skip

PS I'll get some pictures up on my website when it is put to bed. sh

[swt]

You know...i was tempted to ask that sometime. I'm actually using only one cap, 450volts ratings, through the dropping lines. Not the series caps. I don't know how this can change the sound, but i think it changes the impedance of the ps. Maybe someone can help us understand. I was thinking in changing those caps, as per the schem, because opposed to you...i'm getting too much juice and had to use bigger dropping resistors. Let's see if someone more experienced chimes in...

[Luthierwnc]

Well SWT, looks it's just you and me. Maybe I can help you with your question. Those series filter caps and the resistors that run parallel to them help smooth the ripple while converting AC to DC. They are also handy in the first stage to deal with spikes and on-rush current during start-up. Without a rectifier tube, it all slams the system at once.

It also means that there are a lot of ways for B+ to get to ground. With two 330k resistors in series and four sets of 470k/470k going to ground along the string, you are looking at something like 185K to ground from the B+. All those series resistors create five parallel strings (minus the effect of the dropping resistors).

If you took a Blackface Bassman with the same VAC as my transformer, you would get around 445 VDC under the same load. My ODS build comes in around 412 on the plates. My question to the group was, "What is the voltage drop across the filter circuit versus single caps along the line?" What I was hoping to avoid is having to pull the filter board and clip the resistors off the bottom.

I've seen schematics up around the 475 VDC range and Glasman had some builds much higher on his Clone page. What kind of volts are you getting on yours?

Skip

[dogears]

My Glaswerks gets 465v or so....... I have the resistors btw.

Maybe the PS is weak? All 5 of my amps get between 450v and 485v.

[skyboltone]

As a rule of thumb, a capacitor entrance filter will acheive around 1.4 times the RMS voltage of the power transformer. At 350 VAC each side of 0 you should be able to get as high as 490 at the first cap. To get more voltage I would use a full wave rectifier and a large (say 200 mfd) electrolytic as your entrance to the filter. Then you can add the dropping resistors to get what you want at each stage. Be sure to calculate the power dropped across the resistor by measuring the voltage drop across it and deviding it by the resistance. Take the resultant current and multiply by the voltage drop to get the power (wattage rating) of the resistor.

All of this of course assumes a silcon rectifier. Your capacitor or capacitor string should be rated at 1.7 times (min) the half wave voltage of the transformer.

A choke entrance filter will usually produce about .9 times the half wave voltage of the transformer.

With the transformer you are using, there is no doubt you can get the 440 VDC you are looking for.

Dan

[Luthierwnc]

It turns out the diode string was weak. I used 3 UF1007's per side with .01 caps in parallel. I tore them out and used a temporary pair of 1n4007's with all the same filter caps and dropping resistors and got 459 VDC. Go figure. Unfortunately I had to destroy the rectifier caps to get them out so rebuilding will have to wait until I order some more.

Thanks for all and I'll make a more complete report when it is in the box.

Skip

[Luthierwnc]

I keep waiting for the 40 watt bulb to come on over my head.

The trannie is a JCM800 at approx 350-0-350. The power section is identical to the Skyliner. With the power tubes removed, standby on I get 471 VDC at the rectifier. With the tubes in it is 412 VDC. Disconnecting the preamp tubes I get 422 so the power tubes are drawing down 49 volts by themselves. With the screen supply pulled (but volts on the plates and both the B+ and screen supply caps on) I get 465 VDC. Screen resistors are 1k 5w sandboxes. Voltage on pin(s) 5 is -47. Filament volts are a little high at 6.65 VAC

I have tried two different matched sets of power tubes pulling 33 ma each. No red glow, dead quiet, sounds really good for blues or rock. Excellent, actually. A wise man would be happy but I know something isn't right and I have all the Mustang Sally amps I'll ever need already.

Any diagnostic procedures, suggestions or theory would be welcomed

Thanks, Skip

[paddy]

hi there luthierwnc
it seems to me as if the voltage on the plates and screens are
a little to close in value(only 1 volt difference)
the schematic shows a differential of about 7 volts
could something be amiss with the screen wiring?
have you wired in a choke/resistor between plate and screen supply nodes?

[Luthierwnc]

That is what I kept thinking but I had gone over the wiring in detail and everything else worked too well for a component failure.

When in doubt, cheat. I actually built or rebuilt five amps in this run. The other 50 watter has a Matchless laydown PT I got from Steve Melkisethian. 730V CT @ 250ma. Beautiful iron. I bellied that amp up to the ODS, clipped a few wires and jumpered the HV secondaries over to the diode string. 451 VDC with no excuses. Same tubes, same filters, same resistors (although I did rebias the tubes).

The trannie in there is a Mojo Marshall JCM800 50watt replacement. I think Heyboer makes them. Pretty stout piece of iron in its own right. It goes 351 VAC on the hots and drops to 414 VDC under load. It also fluctuates 20 volts on the sweep of the bias pot. The Matchless has 361 on the HV and puts out 451 under the same load.

I guess this will require a re-think. For all the sophistication of the preamp, this is still a hot-rod Bassman, power-wise. I'll have to research new iron. I'd get another of those Angela PT's but they aren't on the site anymore and he isn't the best guy to ask about good deals gone by.

[mlp-mx6]

If yours is a lay-down power xfmr, I'd recommend David Allen www.allenamps.com - GREAT iron, super guy to deal with.

[Luthierwnc]

The laydown is in the other amp. It is easy enough to make a stand-up bracket for the right part. I have used Allen transformers before and agree they are a cut above.

The mistake appears to be entirely mine. A little research on the Plexi Palace and AGA archives suggests the B+ on that amp is all over the map -- depending on the model and year. Anywhere from 400 to 460 is acceptable.

I am still surprised that it loads up so easily. At 33 ma on a really good set of 7581's I am at 421 VDC. That is without the OD on. That setting is about the 1/2 way point of the bias pot sweep. At the hottest setting the volts get down to about 408 and a cold bias will take it up to 433. That is a lot of play. With a set of EL34's at spandex settings it would be below 400 -- probably with a lot of sag.

For now I will probably leave it as is and see what it sounds like outside of the basement shop. With a 7' ceiling it is hard to tell good from great. Bad always sounds bad. I also have a lot of stuff down there that rattles at different frequencies. The cover on the upper wheel of my bandsaw is tuned to low G.

Thanks for everyone's help!

Skip