Matchless DC30 EF86 Channel - too gainy

[greg]

Hello

I have a 15 watt version of the Matchess DC15 which I built a while ago. The EF86 channel sounds pretty cool, but has no headroom. I am sure this is by design and having my build be 15 instead of 30 watts isn't helping.

I am wondering how to increase the headroom of this channel, to allow it to move the point on the volume knob where it starts distorting up to about 12:00 instead of 9:00.

I have attached a schematic - my amp is the same, except that it is using only 2 power tubes and also has a different coupling cap values.

Any help would be greatly appreciated.

Thanks,
Greg

[greg]

The schematic:

I forgot to attach the schematic. Here is is.

Thanks,
Greg

[J Martin]

Adjust the EF86/6267 plate resistor downward. Leave the screen resistor where it is unless you loose too much gain. For lowest distortion the screen grid voltage should be 30% to 40% of the plate voltage. There are other factors, but I find these to be a good starting point. Pentodes are tweaky, but mostly worth the effort.

[greg]

Thanks - I was able to clean up the other channel of this amp by lowering the plate voltage, but it didn't occur to me to try it here, as I expected that the recommendation would be to tweak the screen grid resistors first. I'll give it a shot.

Thanks again,

Greg

[hasserl]

Thanks - I was able to clean up the other channel of this amp by lowering the plate voltage, but it didn't occur to me to try it here, as I expected that the recommendation would be to tweak the screen grid resistors first. I'll give it a shot.

Thanks again,

Greg

Greg, the idea is not to lower the plate voltage, but to reduce the value of the plate load resistor to reduce gain of the stage.

(BTW, don't try to adjust the plate voltage by changing plate load resistors, do it by making adjustments in the power supply. Also reducing plate voltage will not result in increased headroom, it will result in the opposite, the tube will go into non-linear operation earlier. Increased plate voltages will increase the range of linear operation.)

[greg]

Oops - you are correct, I lowered the value of the plate resistor to increase the headroom on the other channel, not the plate voltage - but to be honest, I assumed that I was actually increasing the plate voltage when I did it!

I am often amazed that I have built a small stable of good sounding, reliable amps without really learning what the hell I am doing!

Thanks for the reply, and my wife thanks the members of this board for your continued assistance (as I have not purchased or owned an expensive production amp in 6 years, thanks to you guys).

Greg

[hasserl]

I lowered the value of the plate resistor to increase the headroom on the other channel, not the plate voltage - but to be honest, I assumed that I was actually increasing the plate voltage when I did it!

The net result of a smaller plate load resistor may well have been a higher plate voltage. I'm just saying that if your intent is to increase the plate voltage the place to work on that is in the power supply. The plate load resistors set the gain of the stage, and the impedance. You adjust its value to effect those areas.

[Steve A.]

Hello

I have a 15 watt version of the Matchess DC15 which I built a while ago. The EF86 channel sounds pretty cool, but has no headroom. I am sure this is by design and having my build be 15 instead of 30 watts isn't helping.

I am wondering how to increase the headroom of this channel, to allow it to move the point on the volume knob where it starts distorting up to about 12:00 instead of 9:00.

I have attached a schematic - my amp is the same, except that it is using only 2 power tubes and also has a different coupling cap values.

Any help would be greatly appreciated.

Thanks,
Greg

Greg:

This is totally irrelevant but a buddy of mine has an SC-30. I told him that he could switch between the two channels with an A/B/Y box, but he said that the amp just doesn't work for channel switching because the EF86 high gain channel overpowers the clean channel so much.

So it occurred to me that would be the reason that Mark Sampson is putting the Master Volume on just the Gain channel of his Hot Cat amps (the old "Cut" control is now called "Brilliance" and he reversed the direction of the pot so CW is brighter).

In any case, I figured if I ever built a DC-30 clone (actually probably a 15 watt version like yours!) I would remove the cross-channel master volume control and just wire it to the PI input from the gain channel.

I was wondering what transformers you used for your amp?

Thanks and excuse the irrelevance...

Steve Ahola

P.S. Speaking of which, the Hot Cat 15 has a really simple clean channel: one gain stage going into the volume control going into PI. Period. You set the Clean volume, dial in the tone with the Brilliance control, and then go ahead and set up the Gain channel- which goes through 3 gain stages and a dc coupled pair/cathode follower before getting to the PI.

[greg]

Yep, I used to have an SC30 as well and I never was able to use an A/B box with it for that very reason. It never occurred to me that the master could be put on just the one of the channels.

With this 15 watt version, I definitely have the same problem and I think your recommendation is a great one. I skipped the Master altogether on this one (just employing the Cut).

Being that I am primarily just using the EF86 channel for lead, the volume boost is actually welcome, but if I can get it to obtain less overkill distortion, I expect that it will still have the same difference in volume between the two channels and would benefit from applying the Master volume per your suggestion.

Can you point me toward a schematic of this type of Master Volume circuit? I have always read that Matchless Post PI MV was different than most, but I have never built an amp with a MV, so I've never looked into it.

As for the OT, I built it using the original Hammond L100 OT that was tied to this junk chassis. I did the same thing with my first build (a Spitfire - I owned a real one that that time) and although it sounded pretty good, truly became a Spitfire when I put in the West Labs OT - it totally increased the bandwidth across the board.

Once I determine this DC15 is not going to be scrapped, I will be using a West OT on it as well.

Thanks for the tip.

Greg

[greg]

Question for Steve A

Greg:

This is totally irrelevant but a buddy of mine has an SC-30. I told him that he could switch between the two channels with an A/B/Y box, but he said that the amp just doesn't work for channel switching because the EF86 high gain channel overpowers the clean channel so much.

So it occurred to me that would be the reason that Mark Sampson is putting the Master Volume on just the Gain channel of his Hot Cat amps (the old "Cut" control is now called "Brilliance" and he reversed the direction of the pot so CW is brighter).

In any case, I figured if I ever built a DC-30 clone (actually probably a 15 watt version like yours!) I would remove the cross-channel master volume control and just wire it to the PI input from the gain channel.

I was wondering what transformers you used for your amp?

Thanks and excuse the irrelevance...

Steve Ahola

P.S. Speaking of which, the Hot Cat 15 has a really simple clean channel: one gain stage going into the volume control going into PI. Period. You set the Clean volume, dial in the tone with the Brilliance control, and then go ahead and set up the Gain channel- which goes through 3 gain stages and a dc coupled pair/cathode follower before getting to the PI.

Steve - I am in the process of taking your advice on my little "DC-15". The Lightning side is simply pretty lame - I may have made an error on it, but it isn't doing what I want. I'll likely give the Bad Cat clean channel a try - thanks for the tip.

Also, on the EF86 side: I added a pre-PI MV to it using a 1M log pot. While it does reduce the volume, it doesn't really seem to work like I was hoping (also, please remember, this is the first MV I've put in an amp). First off, it is scratchy. I was assuming that since I put it between the .01 cap and the PI screen, I'd be working with a signal that has no DC on it. I'll check that out soon.

More importantly, however, it seems that the distortion I am getting from that channel is very much a combination of the EF86 and the PI. When my EF86 channel master is down to a level which matches better with the Lightning side, it sounds quite different than with the master up high, sending a lot of signal to the PI. Noticeably less distortion (and not bad sounding, but not the "lead" channel I was hoping for).

I suppose my real question is should I be using a 1meg pot for this application and does it simply need to be inserted between the EF86 stage and where it enters the PI (with the pot tied to ground on one leg) - or is there more to it than that?

Finally, am I experiencing the expected results?

In short, what I have built is a 15 watt boat anchor with a bunch of tubes. Well, it is not that bad, but not was I was hoping for. The Lightning side isn't clean enough at usable volumes (and simply doesn't sound very good), the EF86 channel distorts great, but is too loud. Taming its volume by way of the pre-PI master reduces its distortion too much for my original intended purpose.

Any direction on salvaging the build would be appreciated.

Greg

[DirtyGeorge]

Greg,

The grid of the PI is several 10s of volts above ground. To put the master volume on the PI input you would either need to add a blocking cap or instead of grounding the bottom of the pot, tie it the to the top of the tail resistor.

But putting a volume pot there is worthless anyhow since it just follows the channel volume. So they both do the same thing, except that having the pot tied to ground you are screwing up the PI bias.

The main problem you are facing in trying to have balance between clean and distortion channels is that, as you have already seen, the distortion is being produced in the PI. You need to overdrive the ef86 instead and then your existing channel volume will control the level of the distorted signal.

To do this I would split the parallel triodes in channel 1. Leave one for that channel but use the other to boost channel 2 input before sending it to the ef86. Put a volume pot between the triode stage and the ef86 to control the amount of distortion. Use the original volume pot to match the output level of the ef86 with the clean channel.

DG

[greg]

Additional gain stage before the EF86

I like this idea. The "clean" channel preamp I am about to try uses only 1/2 of the tube, so I'll have one available.

With it set up as you describe, will I be able to wire this volume pot such that it is affectively off for this pre-Ef86 gain stage allowing me to tap into the original signal path to retain the original sound? Or, would I need to have another input jack that takes me straight into the EF86? Better yet, some kind of switched volume pot?

You are leading me places I have never been - this is good. This whole project has been a great learning process (It started out as a Lightning, but I had months of oscillation issues, I replaced the OT, still had oscillation, so I added the EF86 channel just to see if the oscillation was in the Lightning preamp, then a trip to a real tech (which bore no fruit), then I found that I had miswired the OT, - then finally - a working amp that simply hasn't yet been worth the effort (tonally).

This sounds to me like an idea worth pursuing. Definitely not the simple, single channels, non-MV stuff I have built previously.

Thanks for the suggestion.

Greg

[DirtyGeorge]

Greg,

You could use a SPDT switch (or switched pot) to bypass the added triode gain stage. Connect the switch common to the ef86 grid and one switch contact to the input (junction of R5 and R6) and the other to wiper of the first vol pot.

You might want to use a DPDT switch instead. With the second switch circuit connected to the coupling cap from the triode plate. With this you would short the first triode output to ground when in bypass mode. I don't think this would be necessary but keep it in mind.

This should give you the original tone (when in bypass mode) but of course you will still have the PI distortion and the clean channel balance issue.

I am going on vacation tomorrow so I won't be checking the forum for two weeks. Good luck with you build. I will check this thread when I get back.

DG