Gibson EH185 clone

[GHR]

Hi all,

I've posted this question over at AX84, with no luck so I'm going to try here.

Here goes: I built a clone of the Gibson EH185 and the low frequency is a little buzzy. I did not follow the original power supply in that I used SS diodes and I've used series resistance with a normal choke to mimic the original field coil. Voltages are pretty close in the PA but the preamp voltages are about 10-15V high.
I also added screen and 6L6 grid stoppers. 1K for each.

The circuit uses a tone stack employing a choke for the bass control. I had guestimated a 20H choke would work, but had little effect. The stack is similar to the Thordarson Dual Tone control
http://www.theused.com/manuals/thord...ne_control.pdf and I rewired it with a dual pot arangement in order to compensate for the lack of data on which choke to use for the bass. I ended up with a 10H 500R choke which came from some 1940s military gear I got from a hamfest. It almost works and with the thordarson dual pot arrangement I can get full control.

This leads me to the question at hand:

I still have too much buzz in the low frequency response though. I may just be cold biased but I think it may also be the local NFB resistors from the 6L6 plates to the PI plates need adjustment. I'm using a 4.2k OT which may not be giving me enough NFB compared to the original circuit. I have no idea what the original OT load was.

Here's the schematic. I've circled the resistors in question. Note that this schematic uses Roman Numeral M for 1000 instead of K. So 250M is 250000.

EH185.jpg picture by ghramsey - Photobucket

I really would like to adjust them with a pot. I think a dual pot would be better but I can only find cheapo 1/2W types which I'm nervous to put 270V and have it wear out and short.

The existing 250K resistors I have in there have 156V on the PI side and around 270V at the 6L6 side with 108V across it.

So what is the best way to make them adjustable? I don't want to solder in new resistors since I wired these to the socket lugs and they're recycled sockets with fragile lugs.

I have some 500k clarostat pots which I could try but do I need a minimum series resistance between the 6L6 and PI plates? Surely shorting them would be bad for both tubes? What should the minimum be?


Any helpful insights will be appreciated,

Henry

[Steve Conner]

Can you explain what you mean by "the low frequency is a little buzzy"? Is it -

When you play a bass note, you get buzzy sounding distortion along with it
(or)
The amp makes a 60 or 120 cycle buzzing sound whenever it's turned on

Either way, it could be normal behaviour for this circuit.

[Jack Hester]

The circuit uses a tone stack employing a choke for the bass control. I had guestimated a 20H choke would work, but had little effect. The stack is similar to the Thordarson Dual Tone control http://www.theused.com/manuals/thord...ne_control.pdf and I rewired it with a dual pot arangement in order to compensate for the lack of data on which choke to use for the bass. I ended up with a 10H 500R choke which came from some 1940s military gear I got from a hamfest. It almost works and with the thordarson dual pot arrangement I can get full control.

When you say that you rewired with dual pots, does this mean that you now have four pots? I've looked at this tone control before, and never found stacked pots of the values in the schematic.

Also, what do you mean 'almost works'? Thanks.

Jack

[GHR]

When you say that you rewired with dual pots, does this mean that you now have four pots? I've looked at this tone control before, and never found stacked pots of the values in the schematic.

Also, what do you mean 'almost works'? Thanks.

Jack

I originally wired the tone stack exactly per the Gibson schematic, but I used a 20H/1850R choke since I didn't know the actual inductance of the GQ13 inductor in the schematic. Thordarson's Dual Tone Control is nearly identical except for some resistor/cap values and their use of two stacked pots. The extra section of each pot is the same as the 500k/gnd/500k divider between the two controls on the Gibson schematic.

In thordarson's diagram the 'dual tone control' is a stacked 500k/40k custom pot unit. You have to zoom in really close to see the wiring layout. The closeup of the wiring shows the pot values, as 500M and 40M, btw. So they're 500k and 40k.
The parts list just shows them as 'dual tone control' and thordarson's part number.
http://www.theused.com/manuals/thord...ne_control.pdf

Since I could not find thordardson's pots and the only one Ebay had sold before I could bid on it, I took a dual-1M linear and put resistors across it to get 500k and about 40k. Then I rewired the whole tone stack per the diagram in the linked article.

With this rewired arrangment I get good range out of the controls. What I meant by almost works is that I still may not have the correct inductor. The 10H choke makes the bass so loud and overpowering that it rattles my floor, walls and closets doors. So I must not put the control to max bass.

Which leads me to the NFB resistors. I suspect not using the same OT load doesn't give me the correct amount of NFB. This in turn affects the low frequency response. NFB linearizes(sp?) the frequency response so perhaps it will tame the low end response too.

Which is why I would like to vary those 250k resistors from 6L6 to PI plates. I suspect a minimum resistance is necesary though, but how much is what I would like to know. How low can I go? 47k-100k?

[edit] I also want to note that the boomy/buzzy sound is on the mic channel which has much higher gain compared to the instrument channel. I also just lowered the cathode resistor to bias the tubes hotter. I have been mistakenly thinking the 6L6G max disspation is 19W but it's 21W. They're now set to idle at right at 14.8W and 15.1W so they're close to being 70%. The boomy/buzzy bass is not cured though. [end edit]

[GHR]

Can you explain what you mean by "the low frequency is a little buzzy"? Is it -

When you play a bass note, you get buzzy sounding distortion along with it
(or)
The amp makes a 60 or 120 cycle buzzing sound whenever it's turned on

Either way, it could be normal behaviour for this circuit.

60 or 120Hz hum is actually pretty low unless I crank the mic channel. Then I'm getting some EM pickup from the inductor in the tone stack. Hum is acceptable and way less compared to a few of my other builds.

The sound is boomy & buzzy distortion when I play a hard low E note or much of the notes on the low E string from fret 5 down.

There's also a resonance peak on the open-A string and at 5th fret of LowE too, but that may be the inductor or my speaker. If I turn the gain up too much and leave the strings open (just leaving the guitar plugged but not touching the strings) the
A string will begin to vibrate on it's own and then massive feedback occurs. I'm not sure this is related to the issue at hand though.

I think I've forgotten to note this is in the mic channel only. The instrument channel has very little gain and I have to max the volume to hear much from it. This may be normal though.

The mic channel is the issue here. I have to keep the volume pretty low or it just becomes bottom heavy then buzzy.
See my post to Jack Hester above/below too. His question and my reply are somewhat relevent.

I want to change the NFB resistors from the PI plates to the 6L6 plates. A reliable contributor at AX84 said they're local NFB.
If so shouldn't changing them help with some of the boomy, buzzy response? NFB linearizes the frequency response so I should think it may help to change the resistors' value.


Since I'd like to change to a pot for each I suspect I need a minimum resistor between the PI and 6L6. Any ideas what the minimum resistance would be? I don't think shorting the PI and 6L6 plates would be a good thing right?

I'm going to lift them both so there's no NFB tonight and see what I get before I change anything else.

And I may try biasing a little warmer. I make the bias to be about 63-64% of 19W for 6L6. This is perhaps too cold.
I'm going to see if getting to 70% has any effect. [edit]I lowered the cathode resistor to 150R and the tubes are now biased right at 70%. I've been erroreously thinking the 6L6G max plate dissipation is 19W but I just realised it's 21W so I was biased colder than I thought. However, the buzzy/boomy low end persists so it was not caused by cold bias.
[end edit]

[Steve Conner]

That's probably just the way that circuit sounds when overdriven. You may need to modify ("revoice") it to cope with modern playing styles and/or humbucker guitars. There was recently a thread about "what's your favourite 5E3 mod" and the advice given there may be relevant. Mainly making the interstage coupling caps smaller to filter out some bass.

You can probably remove those local NFB resistors altogether, too. (A 1M dual gang pot with the original 250k in series would be a good adjustable replacement.) Having said that, their main effect will be to damp your speaker, so if it has a heavy bass resonance, it'll only get worse.

[GHR]

That's probably just the way that circuit sounds when overdriven. You may need to modify ("revoice") it to cope with modern playing styles and/or humbucker guitars. There was recently a thread about "what's your favourite 5E3 mod" and the advice given there may be relevant. Mainly making the interstage coupling caps smaller to filter out some bass.

You can probably remove those local NFB resistors altogether, too. (A 1M dual gang pot with the original 250k in series would be a good adjustable replacement.) Having said that, their main effect will be to damp your speaker, so if it has a heavy bass resonance, it'll only get worse.

Thing is it may not be just how it sounds. There's a youtube clip and the person has a real EH185. But I think it's the later one with a the single tone control. It's probably not a good example but it doesn't seem to have much buzzy or boomy-ness.

Maybe my speaker just doesn't have good low frequency response. It tends to not like large low end signals. It's a 10" celestion v10-40 btw. I cannot find much information on it since I believe it's discontinued. I got it when I was just getting into tube amps back in 2000 'cause it was about $70.
If I had to do it over I might have bought something else.

Would more NFB help with that? Can I reduce the 250k NFB resistors to 100k or so and use a dual 1M pot to make the NFB adjustable? Reducing the 0.1uf coupling caps will reduce the bass that hits the 6L6s. Maybe I should consider that? I built it pretty stock and I've been modifying it ever since.

Can you post a link to the 5E3 thread? I did not locate it with a forum search. It found only two hits: this thread and one about a vibroclone. I'm going to search again so I'll probably find it by the time you reply.

[Steve Conner]

http://music-electronics-forum.com/t28168/ voila!

[GHR]

Thanks for the link. For the most part I'll probably end up swapping coupling caps.
Before I try that though, I tried my idea to reduce the NFB resistors.

I swapped in a pair of 100k/1W fixed resistors (one for each PI/6L6 plate) in series with a dual-500k linear pot.
I powered through a 100W bulb in case the 100k minimum wasn't enough. Even accounting for the underbiased condition (the 100W lightbulb took the voltage down to 200V plate and only 11.8V cathode so it was rendered cold bias) the sound was much improved. The extra feedback absolutly improves the low frequency sound.

Even with the bass control to it's max bass position which usually gives over powering bass it was acceptable.

Cranking the mic channel volume still overdrives to a buzz, but that I understand and I probably won't do since 18W this amp is rated to is way more than I need. I can now get to at least 5 or 6 whereas before it was buzzy at may 2 or 3.
I'm just not sure about this pot's ability to handle DC across it for long. It's old I think. I was beginning to smell an odor like something cooking so I didn't keep it powered on for long. But when I checked for heat neither the 100k resistors nor the pot was even the slightest bit warm. Perhaps I'm being paranoid about the wattage?

Either way I may just sub 100k resistors in place of the 250ks for now. This exercise certainly proved that I can go lower there and make it sound better. I think I could even go down to 47k, but I suspect I need to go to at least 1W resistors. The 250k resistors were dropping 108V. That's 0.000432A and thusly 0.047W. If I go to a 47k resistors I think I need at least 1W resistors. I bought some 47k-1/2W resistors. Without knowing for sure they're not going to drop too much voltage and burn up I am almost afraid to try them. Any thoughts on this?

[Steve Conner]

If it were me, I'd be happier with 1W. The little half watt resistors have a maximum voltage rating too, 350V or so, and there's a lot of AC voltage on the power tube plates.

It might be worth making the PI plate resistors bigger too. That would help to restore the DC conditions and also increase the feedback a bit more.

Too much feedback will probably make it oscillate, but how much is too much, I have no idea. Probably if you left out the PI plate resistors altogether, that would be too much!

[GHR]

If it were me, I'd be happier with 1W. The little half watt resistors have a maximum voltage rating too, 350V or so, and there's a lot of AC voltage on the power tube plates.

It might be worth making the PI plate resistors bigger too. That would help to restore the DC conditions and also increase the feedback a bit more.

Too much feedback will probably make it oscillate, but how much is too much, I have no idea. Probably if you left out the PI plate resistors altogether, that would be too much!

I put in 47k with the dual pot and w/ the control at minimum so only the fixed resistor is working the voltage drop is only 80V across the whole works. Since the pot is at min this means it can be ignored and the 80V is effectively across the 47k resistor. No problems so far. Though I agree about using a higher wattage resistor. I'll see to that this week coming.

The sound improved even more the lower resistance I've gone and I think it's also reduced the gain a little.
The pot is almost unneeded since I end up at it's minimum travel so only the fixed resistor is doing the work.

I've also discovered that there are some series, parallel or both resistance interactions going on between the NFB resistors and the two plate circuits. Something is making the resistance appear lower when in circuit.

When I read the resistance across the 47k resistors and the pot together I get only 80k. If I lift the connection from the pot to the PI it goes to 470k or so which is right for these pots. This explains the lower voltage drop than I expected I might have. The 47k is being bumped up to 80k by these parallel/series resistances. Care to hazard a guess where I'm seeing the extra resistances from?

Also, if the PI is 158VDC and the 6L6 plate is at 260VDC (I lowered the cathode resistor again), shouldn't the max voltage drop the resistor can see should be the difference between the two: 102VDC?

If I have sufficient resistance it shouldn't ever drop the whole 102V since that would short PI to 6L6.
102V/47000=~2.1mA
I make that under 1/2W of power.

So perhaps that 1/2W resistor would survive though that's not accounting for the AC riding on the plates of both tubes which is probably substantial. I am going to measure that again and see.
I think it's a few hundred volts though on hard signals. One more argument for higher wattage.
I'll have to get some 1W or prefereably 2W-3W flameproof resistors. But they won't be here until end of week
even if I order today.

So far I have had zero oscillations btw. I've messed with variable NFB in an AX84 circuit with a 6SN7 PP triode output and it motorboated when I had too much NFB. This one just seems to make the bottom end sound cleaner the more NFB I get.

I'm tempted to go to 22k (of which I may have some 1W or more) and see how low I can go.

With the lightbulb limiter at least I'll know if I've reached the low limit of resistance between the two plates.

I'll look to change the PI plate resistors at some point, but not just yet. Due to my use of eyelets they're more difficult to change since I used short wires and eyelets have minimal mechanical connection. I've already rewired the twice PI due to being the wrong phase compared to the OT so it squealed. Now the wires trimmed so short if they break I'd have to use new. I'd prefer not to have to do that.