Upgrading the VHT "WATTS" Circuit

[Looselectron]

Hi! They call me Adam Smith. What do I do to an amp that has a MOSFET voltage controller for its power tubes that FAILED TWICE?? 

First of all, I ordered a replacement. At the same time, I ordered the SV2+VCK kit for my VHT Special 12/20RT head but the instructions say make sure it's working properly without power scaling. I think I have the amp almost working without the original voltage starvation circuit but need to compare the test voltages against my second VHT Special 12/20RT combo (for a stereo rig) that still works...for now. 
   

It occurred to me to try installing the VCK first in the working combo to see if it makes any difference with the existing WATTS circuit before the SV2 in the head I have on the bench. Interestingly these amps have a RETURN control on the effects loop which effectively works like a master volume so I think I can safely omit the Drive Compensation control for this installation.     

Updates to follow...

[K O'Connor]

Hi Guys

You can see in the VHT circuit that the range for power reduction has been severely restricted inasmuch as the minimum power level equates to a still-too-high loudness. R65 needs to be reduced to 33k to allow useful sweep.

The SV2 we supplied has a tweaked resistor value to allow the use of the 1M dual pot in the VHT, as the other section is used to proportion the tremolo modulation signal applied to the output tube grids.

The amp lacks any means to control the rest of the circuit voltages to keep their tonal contribution consistent, which might be why they limited the amount of power reduction possible. This tends to be the way, as well, when combo-format amps are given poorly thought out power reduction schemes. Even Mavin Peal did that with their combos because the speaker loses fullness of tone at low-loudness, but also MP scaled too much circuitry. The VHT scheme seems influenced by Dana Hall and his vvt, which was a copy of a portion of the Classic-PS approach.

Physical construction of VHT amps we have seen is pretty tight, and sometimes opaque as to how to take them apart. Thermal management is therefore a bit compromised even for the tubes, let alone for fragile, heat-sensitive mosfets. This is the main weakness of power mosfets, is their need for very good cooling and heat extraction, and would explain why the original mosfet Q1 has failed twice. Mounting SV2 in the same place won't be more reliable.

In a combo amp, adding a fan is a good idea and fan noise will not be a problem. Use a 12Vdc fan and power it from 8Vdc derived from the 6Vac heater supply. A bridge and filter cap of 2,200uF minimum will suffice. The fan turns at reduced speed and you will not hear it at all from in front of the amp. The fan should be in the cold air path and push air past the tubes towards the rear. This is the same for stand-alone amps as heads and is shown in TUT and subsequent volumes.

Have fun

[makinrose]

I've also worked on some of these VHT amps and they do leave something to be desired. Installing the SV2 will be a big improvement over the way the "scaling" is implemented now without a current limiter.

Occasionally, what I've done on amps that I can fit a fan in, or the player won't accept one is to mount the mosfets in a diecast aluminum box attached by an umbilical cord like some old organs and other tube equipment that had multiple chassis used to have. Since the box mounted on the bottom the heat rises out of it and hopefully if the combo is large enough it won't be affected by the heat of the tubes.

[K O'Connor]

Hi Guys

Mosfets have a very wide frequency response and are prone to oscillation if not treated carefully. They must have a gate-stop resistor right at the gate, just as at ube should have its grid-stop right at the grid and a BJT should have a base-stop right at its base.

Some mosfets have built-in gate voltage clamps in the form of zener diodes, but not all do. Those zeners are typocally 20-30V and if they fail due to an excessive input current at and excessive voltage, then the entire device is toast - burnt toast. This is why you see a zener diode across the G to S terminals, or including the gate-stop.

The London Power kits that use mosfets have the gate-stop and gate protection zener right at the mosfet mounting pads. If you choose not to mount the mosfets on the board these few other parts must be relocated with the mosfets to keep the mosfet from oscillating.

Wiring to the "protected mosfet" should be twisted - do not use coax.

The obvious alternative is to leave the mosfets on the PCB and move the entire unit to where it is cool. This means running twisted pairs of wires to the various wire connection blocks. We offered this as the PS-BOX many years ago, which was during the Classic-PS era and had a few issues. Many of those issues are eliminated with the current kits, but it is still preferred to have all the Power Scale circuitry inside the amp chassis.

Note that the external box must be tied to chassis ground to assure that it does its task as an EMI shield.

[makinrose]

I should have mentioned that when doing this I only locate the mosfets and the Zener's in the box so all other circuitry remained inside the amp . I generally use the older DC PS circuit with the 2W Military type pot for the PS control . My experience is limited to that type of installation. I haven't had any issues though.

[K O'Connor]

Hi Guys

The gate-stop must be there, as well. The SB-style circuit had that. The resistor helps protect against oscillation. The zener protects against voltage spikes.

The frequency of oscillation can be in the tens or hundreds of MHz and your scope may not show it but the mosfet will overheat and likely die.

[makinrose]

Hi Guys

The gate-stop must be there, as well. The SB-style circuit had that. The resistor helps protect against oscillation. The zener protects against voltage spikes.

The frequency of oscillation can be in the tens or hundreds of MHz and your scope may not show it but the mosfet will overheat and likely die.

Good to know!

[Looselectron]

The SV2 we supplied has a tweaked resistor value to allow the use of the 1M dual pot in the VHT, as the other section is used to proportion the tremolo modulation signal applied to the output tube grids.

Step ZERO is getting the amp to work without WATTS, and "Assure the amp works properly before doing the SV2 + VCK installation." 

Unfortunately there are no clear instructions for this. So I started by yanking out the unnecessary parts.    

Before diddling around with the most lethal part of the circuit, I contacted VHT about the removing it. They sent me the following image but those extra diodes were already on the board! This is for the Special 12/20 (sans RT) so R56=69(B+3), R59=67 (B+1) and R55=64 on the Special 12/20RT schematic.     

So I jumped jump D13/R64 to R67 (B+) and the OT center tap to C49 (D14/B+1). R65 and R66 are now omitted. I hope the additional diode in the supply line compensates for the absence of the VC MOSFET between B+ and B+1 but is a 0.7V  drop still necessary here? Note the daughter board in the upper right corner erected like a car bonnet.     

That left a brown wire from TREMOLO (left lug) and a RED wire leading to TREMOLO OUT on the daughter board. I wanted to splice these but the brown wire just fell off... Jumping TREMOLO OUT to the left lug of the TREMOLO pot was the right move anyway. There was one more stray red wire I had to connect from the reverb transformer to B+2 but the reverb buzzes now even after reorienting the chassis...this has always been a confusing part of the amp for me but I do like the sound of this reverb.     

The amp sounds fine without reverb but until I can get that working 'properly' again, I'm gonna install the VCK separately into my newer VHT Special 12/20RT combo that hasn't failed...YET. I never fully understood how important heat management is in electronics until I started this project. Instead of drilling a hole in the chassis, I took apart a broken induction heater to look for a heatsink. This one might be overkill, but there are probably a lot of other good parts for guitar amps in there including a FAN.       

More updates to follow...

[K O'Connor]

Hi Guys

Looselectron: I believe you went a little overboard with respect to checking that the amp works prior to installing Power Scaling.

In this amp, since you have an issue with the stock mosfet dying, the simple test wiring would be to remove the mosfet, then jumper across where D and S of the mosfet used to go. Done.

You can still twiddle the Watts pot and have the tremolo depth vary and this won't hurt anything.

As far as heat-sinking goes, usually VCK only carries the 12A_7 currents, so bolting to the chassis is sufficient. In this amp, there is a 6V6 as the reverb driver BUT it is powered from B+2, which is also the screen node for the output stage. We do not have to worry about this with respect to VCK.

It is the SV2 that needs big cooling, particularly Q3 which handles plate current for the output stage. Note that Q4 will pass the screen current and the reverb driver current, the latter which is only 5mA max.

Regarding the VHT-supplied mod: The diodes either side of the mosfet are there to protect the mosfet from capacitor discharge currents during turn-off. The input cap may drain faster than the output cap, as the B+1 and B+2 filters have no bleeder resistors where the input cap does. Add a bleeder on B+2. While you are at it, add a bleeder across B+5.

In the stock arrangement, it does not matter to the mosfet if the input cap discharges faster than the output caps do, as the mosfet has a reverse diode from D to S with the same voltage and current rating as the mosfet channel.

[Looselectron]

Thanks, KOC! If I went overboard, it's because I'm learning to scuba dive AND trying to cure amp cancer at the same time. The amp went into a coma again because I confused the two red wires for the OT and RT (reverb transformer...)

After almost ten years of (ab)use, I found more than a few resistors charing and flaking in there, so I think it's probably a good time to do an overhaul.

With respect to the VCK, I am planning to insert it before B+3 where it should also "clamp" voltage to B+4 and 5 downstream as well, right?

[K O'Connor]

Hi Guys

Yes, that is the correct place for VCK in this amp.

Have fun

[Looselectron]

Cool. So I finished assembling the VCK, and prepped the VHT combo for insertion between R32 and B+3 to V5 (phase splitter/tremolo.) 

All I had to do was just clip the red wire below B+3 on the daughter board, and splice in the VCK's input/output!    

The test voltages I got were not what I expected: 

B+3 is about 320V without the VCK but I could only get it up to 240V between the input and ground of the VCK with the trimpot. 

I'm not sure I did this part right but it works!

Rolling down WATTS leaves the amp with more clarity with less of the harsh pick attack it had before. The overdrive is more nuanced and less squishy now at the cost of some unnecessary saturation. It’s like I traded a germanium fuzz for a treble-booster which I know doesn't exactly sound like an improvement to to most but I really like playing with both. 

It might be cool to switch the VCK in/out of the parent circuit as an option to keep the stock sound of WATTS. Updates to follow on the SV2!

[K O'Connor]

Hi Guys

As the VCK notes say, you have to measure the voltages on the amp at the normal idle condition. For a conventional cathode-biased amp, this is more or less the same as the full output condition. In this amp with its power control, the power must be set for full voltage so the output tubes idle as hot as they are supposed to.

VCK is clamped internally to a 200V maximum voltage difference between input and output to protect the devices onboard. This is more this the usual regulation voltage swing found in most amps. The VHT 1220 has a hi/lo voltage/power setting, so adjustment of VCK may be a little trickier and it may end up optimised for one switch position over the other - installer's choice.

When stuffing VCK, the first transistor you install should be the BC, then install the A92, then the A42/3.

[Looselectron]

Last night I affixed the VCK inside the amp chassis with velcro tape, and toted it to a rehearsal. Twelve watts is just barely enough to keep up with a drummer and a bass amp taller than the bassist so there was no need to adjust WATTS but the tremolo depth was too weak to be useable for psychedelic garage rock.

When I initially tried to calibrate the VCK, I noticed that the amp was set to HI (420V B+ for 20 watts) instead of the stock setting but I have always preferred this with 6L6GC pair for a more articulate and dynamic guitar sound.

I think this time I'll just try recalibrating it to get the tremolo back to normal if switching to HI doesn't solve the problem for me.

[K O'Connor]

Hi Guys

As I suggested, calibrating VCK at the high voltage will provide optimal performance at that setting, but switching to quite different input voltage will require a different adjustment of VCK.

Your idea to adjust VCK to where the trem works properly at the LOW setting should provide a compromise for the two settings.

Have fun