Ampeg SVT

[CJ]

working on a vintage SVT,  6 ea.6550's,  700 plate volts (yikes!)  350 watts of heavenly bass,

fork lift required,

 

[Gus]

I worked on one once. Too heavy and a pain work on. You can hurt yourself due to the mass.

 

[CJ]

now you tell me!  ;D

preamp board,  12DW7 dis-similar triode drives 2ea 12BH7,  12AU7a will work ok here as the grid loads of the 6550's are pretty light,

they had a 12AX7a in there for the DW7 which is not good as  the second section of the DW7 has an amp factor of 17 instead of 100like the ax7,  this is a sensitive beast that is capable of self destruction so you want to set it up right.

 

[CJ]

we do not like the filter cap  arrangement,kind of kludged in there like an after thought, loose tie wraps, sitting right on top of 700 volts, if the sharp cap terminals of the can caps punch thru the upper caps, then there might be a power outage somewhere in the neighborhood as this thing uses a 10amp fuse,

 

[CJ]

we punched the chassis for a 100uf/500 and fabbed a plexi termplate (upper rt) for the second cap in the B+ side that feeds the screens, the original was a 30uf at 600, since not many people make a 30/600 we decided to split the voltage load into 2,

this amp has solid state rectifiers, so when you first turn it on, and neglect to use the standy by sw (which controls the B+ transformer) then you could have 730 volts on that 600 volt vintage cap,

now we have a safe power supply and lots of work room,

and why not change the original tube sockets which have been tortured by heat and vibration ad excess voltage with some new ceramic sockets, so we do not loose the bias on one of the tubes  due to a loose socket which is probably why this amp broke in the first place,

 

[CJ]

here is the pwr supply showing the cap we upgraded>

 

[CJ]

we did not care for the high voltage spk jack switch mechanism from the civil war so that was the next mod,

also check the big spk wires from the OPT getting necked down to 20 ga wire,

 

[CJ]

so we use direct wiring into switchcraft 12b jacks (stereo) and wire the extra jack terminal to the ground terminal to get a tighter jack with more contact area for  gnd, kind of like a poor man's speak-on cable system, we want good damping fora bass amp,

 

[CJ]

love the heavy duty tube clamps from the ham radio guys, this amp uses 10 ohm plate resistors, 5 watt, and 22 ohm 2watt screen resistors, and bias is achieved by 2 pots, one for each bank of three pwr tubes, so you want those 2 sandblock 1 ohm resistors in the middle to be matched, we took them out and put them in series with a dc pwr supply and measured milli volts across each one, since we lack a precision ohm meter,

after finding that one was off by 10%, we replaced them with 2ea. matched 10 ohm 3 watt resistors which hopefully will not burn up under full power, this gives us a reading of 72ma x 10 = 720mv across each (cold biased amp) so we get our voltmeter above the noise floor, this might also help keep one of the tubes from drifting into thermal meltdown, which is the biggest complaint on these amps,

we also clamped down the monster pwr cable with a nylon cable clamp

 

[CJ]

here is the amp as it came into the shop,one of the pwr tubes was fried but the rest test strong so we found a 6550 in the pile that matched up and the amp runs fine now,

there is a fan on the back grill which makes working on the amp like working on an internal combustion engine as it sucks in a lot of excess baggage that we do not want to claim,

 

[CJ]

here is the pwr schemo, this amp has a cool pot for matching ac signal into the tubes (blue) so the load is balanced, it could be scaled a little less course, engineers always go for max headroom when it comes to trim pots,

bias pots in red

notice the weird stand by schem, there is a pwr xfmr for the heaters which is as mart move, (see mesa triple rect thread)

 

[CJ]

went thru the preamp and everything was fine, we did replace the filter cap and upgraded the dropping resistor which had the color code  cooked off due to heat, (blue)

got a 10wat ceramic in there now,  and a few more uf on the filter cap,

 

[Wildfara]

Never worked on one of these monsters but I have been up inside my VT40 and VT22 plenty tho. Fantastic amps...don't listen to the naysayers who hate them and call them unreliable. They were built well and die because people don't maintain them (because they were built solid). At the very least they usually need recapped for the main filters the bias supply should definitely be redone. Especially the big capacitor that comes right off the B+ that it is derived from...when that blows your tubes go with it. And the Plate and screen resistors should be replaced.

 

[CJ]

yeah the SVT was like a V4 or V22 with two more tubes,  V4's make great bass amps also,and weigh about the same, 

 

[jensenmann]

I still have one of these, the exact same model like above. These were the second series of SVTs. The first series had blue letters on the front. I used to tour with this one, cased of course, plus the original old 8x10". Lowend heaven but hell for the back. I still remember the day when my drummer went early off to the gig and I had to load that stack alone into my car. OMFG  :'( Glad that my touring days are over...

Smart move with those plate resistors. I remember having upped them in wattage. If one fails chances are high that the poweramp goes south.

 

[ricothetroll]

Hi !

Seems like there's some deep SVT knowledge here 

I'm currently repairing my Ampeg SVT-VR. The schematic is here :
http://www.bustedgear.com/images/schematics/Ampeg_SVT-VR.pdf

It had blown a 6550 output tube so I decided to go for a new sextet. I bought the EHX ones as they are the only ones I could find that are rated for 800V plate voltage, instead of 600V for most other 6550 out there.

Could be interresting to note that the blown tubes has its heater that is open. There are some tiny pieces of glass in the bulb so I guess the bulb is broken somewhere and so the heater burnt because of air (oxygen) got inside the tube.

So I first removed the tubes and checked that the preamp out is fine, and it is. Then I scoped the signal at the 6550 grids  without them installed, and it's fine, I have a clean sine in phase on the 3 "push" tubes and with reversed phase on the 3 "pull" tubes.

Ripple on B+ is 5Vpp, scoped J11 and J12 after the OT.

So I installed the 6550 and biased them at 21-25mA per tube. A 4R/1200W dummy load is connected to the output. Tube bias has been set with TAD BiasMaster, that measures the bias by inserting a 1R resistor in series with the cathode and measuring the DC voltage across it to find the current.

Ripple the is about 15Vpp on B+, which seems pretty OK according to my simulations (Graetz bridge + 50uF reservoir).

When I put a 1kHz signal at the input I get the signal seen on the image linked below : 1kHz modulated by some kind of 100Hz signal. I guess that is the ripple playing...

https://drive.google.com/open?id=136-yKlwcxI9EEcLI9AgI9thDiIgTk7hc

One strange thing that I noticed is that the bias LED circuit doesn't seem to be working fine : LEDs should light separately when the bias of one side is between 3*21.6mA and 3*26mA according to the comparator circuit, but those always light both at the same time and for much bigger currents (like 40mA IIRC)

Another strange thing is that when I put any of the 6 tubes from the new setxet in my tube tester, there seems to be some kind of leak between G2 and the cathode and between G2 and G1, but that only happens when the tubes are fully heated : must wait at least 1 minute before the "leak" LED of the tester starts lighting. This happens for EVERY tubes of the sextet, and when testing them for Gm they seem to work correctly. The "leak" schematic is linked below. It's done by connecting the electrode to be tested to R1-R2, with every other electrodes shorted to ground. Heaters are fed by an isolated supply (trafo secondary). Current should start flowing in the transistor from a leak of 62M to the ground which is quite a small leak IMHO, so I don't think the tubes are dead (hope not actually...)

https://drive.google.com/open?id=1ekVKqMrb6bQfTRsH2I_JXUwvGeTJ-bQh

Any kind of help would be appreciated...

Thanx in advance !

Best regards.

Eric

 

[ricothetroll]

BTW I took one tube that's displayed "leaky" and set a multimeter in G2 circuit to measure screen current along with anode current and I had some pretty normal values :
Ua = 250V
Ug1 = 250V
Ug1 = -15V
Ia = 130 mA
Ig1 = 12mA
gm = 9.1 mA/V

 

[Tubetec]

Top class job as usual Cj.
I never beheld one of the Mk1 or Mk2 Svt's ,did have a couple of the newer models in alright ,even more horrible to work on .The later  models with the led bias were problematic too due to drift , I did read somewhere about a mod which bypassed the led/op amp bias ,but never took the bull by the horns so to speak . Whats your own personal preference on mods for later models ?
Those clamping tube holders look the bomb  too,like they always belonged  there . Uprating the fan to a wide bore 240 volt and providing an on/off switch for studio usage was another option Ive installed too. 

 

[ricothetroll]

I tried to plug a (cheap) bass amp speaker to make sure the problem isn't realted to my resistive dummy load.
The sound I get is just as if the guitar was "ring modulated" by 50 or 100Hz...
Is there a specific conclusion to be made from that observation (hearing) ?

 

[PRR]

You've got (rectified) AC where you need DC.

This is your "ring modulator"; also may be your "leakage" (stray capacitance with AC-ish voltage will "leak").