Vacuum Tube Harp Amps!

[Bruce / Mission Amps]

I've built a bunch, I'm sure others have too.
Let's get some threads running here as most stock guitar amps are not that great for a pure greasy, papery, down and dirty harp amp.

[Chris / CMW amps]

Ok!

About >98% of my customers do play guitar but I do voice amps especially for harp when requested.
As probably mentioned before on various website and Ampage;
- lower voltages
- big coupling- and cathodebypass-caps
- (sometimes) a higher input-Z

Anyone got some suggestions for modding reïssue Bassman amps (stock speakers) for harp? I've done a couple of blackfaces and tweed deluxes but I'm looking for some harp-ideas for "this beast"

tia!

Chris

[Bruce / Mission Amps]

Lower voltages and lower gain in the first and second stages.
Most of my pro player customers like no more then a 12AY7 in the first two stages and some like the 12AU7 with 33K plate load resistors and 510 to 820 ohm cathode resistor.
Another good mod is to use a 12DW7, where one channel is the low low gain triode and the other triode is like the 12AX7.
But on the the high gain triode use a split load pair of plate resistors.
Two 56K resistors and the big .1uF to .22uF coupling cap coming from between the two.
And of course very light NFB.

[jaysg]

Anyone got some suggestions for modding reïssue Bassman amps (stock speakers) for harp?I've never done any harp amps, however, I'm an archivist. Some of these changes may help with a Bassman. I think they're intended for a BF circuit. From one of the former :mad: good guys of Ampage:

1) replace the 1M on the 1st input jack with a 5M resistor
2) pick a channel and replace that channel's 100k plate resistor with a 470k
3) replace the 1k5 cathode resistor with a 2k7 (the 25uF/25V cap stays in place)
4) decouple that stage with a 0.047uF cap
5) decrease the slope resistor from 100k to 56k
6) replace the phase inverter tube with a 12ax7
7) if there's a tube rectifier, use a 5u4
8) replace 6L6's with 6v6's
9) install a dropping resistor to lower the screen voltage: 500 ohm, 10W
10) bias at 27mA/tube

[Chris / CMW amps]

Hi Jay

I've used the tweaks you mentioned in the past in some bf Fenders (together with some other needed "custom" tweaks) and it does work very well. Fat harpsound and less feedback, it's a good starting-point for such amps imho.

Over here it seems most harp players do use those reïssue tweed Bassman amps and I do need to design a harp mod for it; I'm not asking for a 100% ready/tested circuit (altough... :D ) but some good starting points to start tweaking.

As mentioned before: the posted notes for bf-amps are pretty good.

Chris

[MWJB]

It's not what many would expect, but I usually find I end up with around 500v B+ on 5F6As set up for harp. I nearly always end up with solid state rectification, occasionally 5AR4. Run the Sovtek 5881 & fit a bias trim pot (10K & 6.8K fixed resistor) in place of the 22K bias supply dropper, try 8mA upwards.

Double the filtering at the screen supply.

As Bruce suggests, dropping preamp voltages is a good move, it also prevents the amp from being too harsh/spanky with the tighter rectification. Typically, I start with 82K-100K as the preamp dropping resistor.

As for preamp tubes, that'll depend to a large degree on what mic is being used, I always prefer a 12AX7 in the PI. You don't want to drop gain too far or low end & dynamics suffer. The EH 6072A is a killer.

The 6.8K tail resistor in the PI is too small, try 10K or 22K.

Larger preamp coupling caps @ .1uf.

I tend to leave the tone stack values stock, but remove the bright cap from the bright channel. Then you can play with V1B...grid leak bias, differing impedance loads for different mics (most like 5Meg +, except 520DX), 12DW7 in V1 etc...?

Doubling the size of the 220K grid return resistors will make the amp a little more touch sensitive & thicker.

Switch in/out a second 220uf cap at V2, 3?

[Bruce / Mission Amps]

Most of my serious harp player customers laugh at the popularity of the 5-12watt guitar amps used for harp and if they're playing out, always use at least a 30-50 watt amp.
My 3x10 5E7 Harp-Bandmaster with cathode biasing and a 50 watt Super Reverb OT is VERY popular.

Bruce

[Watto]

I've never done any harp amps, however, I'm an archivist. Some of these changes may help with a Bassman. I think they're intended for a BF circuit. From one of the former :mad: good guys of Ampage:

1) replace the 1M on the 1st input jack with a 5M resistor
2) pick a channel and replace that channel's 100k plate resistor with a 470k
3) replace the 1k5 cathode resistor with a 2k7 (the 25uF/25V cap stays in place)
4) decouple that stage with a 0.047uF cap
5) decrease the slope resistor from 100k to 56k
6) replace the phase inverter tube with a 12ax7
7) if there's a tube rectifier, use a 5u4
8) replace 6L6's with 6v6's
9) install a dropping resistor to lower the screen voltage: 500 ohm, 10W
10) bias at 27mA/tube


Can any of these things help with a 5e3 deluxe.I want to tweek one for Harp,
Cheers
Watto

[MWJB]

I'm not crazy about 5E3s for harp, points 1 & 10 can certainly be utilized on a 5E3. You might find that you don't neeed the 6V6s to run as much current for harp applications. Keep an eye on your B+ voltage when raising cathode resistor value, with respect to tube & cap voltage limits.

Other areas I would look at are:

100-220uf cathode bypass caps in the preamp, perhaps no bypass cap on V2 at all?

Rewire pots as voltage dividers or Bruce's tone/vol mod.

I'd think about adding a negative feedback loop (either 6G2 or SF Champ style), experiment with feedback resistor values, you might find that smaller values (more feedback) than found in the guitar amps work well?

Try all the preamp tube swaps you can get your hands on, but don't rule out the possibility of running 2x12AX7.

Speaker selection will play a big part, you need a speaker that is not too lively with regards to feedback, but is efficient with good low end. Many of the more popular 12" guitar speakers may not be such a good choice for harp. The best 12" speakers I have heard for harp have been rather flat sounding when used for guitar (the stock Eminences found in Blues Jrs and Blues Deluxes, Alpha 12?).

Increase the value of the 22K preamp dropping resistor and listen to a few different voltages, say 90-150v on a 12AX7 (then try tube subs, adjust voltage and repeat).

If your PT will take the extra heater current, maybe try 6L6 or KT66?

Never tried to mod a 5E3 for harp, it doesn't strike me as the basis of a good harp amp, so suggestions here are just offered in the spirit of experimentation.

[Bruce / Mission Amps]

1) replace the 1M on the 1st input jack with a 5M resistor

Maybe OK,

BUT, if you put a .05uF cap in front of the grid load resistor, and then get the grid resistor up real high +4M7, that preamp will try and run in what is called contact bias... that does sound different and a little fluffier to me.

2) pick a channel and replace that channel's 100k plate resistor with a 470k

I'd say no

3) replace the 1k5 cathode resistor with a 2k7 (the 25uF/25V cap stays in place)

I'd say no

4) decouple that stage with a 0.047uF cap

I'd say no

5) decrease the slope resistor from 100k to 56k

There is none

6) replace the phase inverter tube with a 12ax7

The PI already uses one of the 12AX7 triodes for push pull adn that section has ZERO gain.

7) if there's a tube rectifier, use a 5u4

I'd say no

8) replace 6L6's with 6v6's

The amp already uses 6V6s.

9) install a dropping resistor to lower the screen voltage: 500 ohm, 10W

The voltage node that feeds the screen's their voltage , with respect to the plate supply is already 5000 ohms and only 2 watts or so!! Why would you need a 10 watter even if you did use 500 ohms.... somebody not using any math.
Seperate screen resistors of 2K7@2W or even higher resistance will give the PA a little more compressed sound. Maybe OK

10) bias at 27mA/tube

I'd say no...
and why pick an odd number like 27ma? That has nothing to do with what the tube is doing unless you know the tube and the B+ voltage, so don't follow any rules regarding ma of idle current without checking the other parameters.


Why to I say no to:
2,3,4,7 and 10

2., The only reason players do this is because many tweakers do not understand what the plate load resistor does.
If there is enough plate voltage delivered to the tube, a larger one can actually increase the gain.
Loosely put, gain is a function of absolute voltage drop across the impedance of the tube, and the ratio of the cathode resistance to the plate resistance, but using a way larger plate load resistor does lower the B+ delivered to the preamp tube which can make it less gainy because there isn't any B+ to work with but it will clip poorly, faster and with less tone too...
that is the wrong way to do it if you are looking for less gain.
A dead stock 5E3 has only about 105v-125v on the plate in the first place so lowering it below 85v-90v is really just silly.

3., That will bias the tube colder and can make the preamp slightly harsher sounding with more odd harmonics then if the tube is biased right in the middle of it's class A space with higher current.... yes it will have a wee bit less gain because of the lack of B+, but there other ways to get that without this method which changes the warm tone of a free running first preamp.

4., Lower impedances like to have larger coupling caps to fully pass the audio at lower frequencies so this is ass backwards.

7., Does not apply to a 5E3 with 6V6s as the amp is a lower voltage B+ amp and the 5U4 will be so lightly loaded with this topology that the B+ will be much higher... ass backwards again.

10., (Except to tweak for personal taste or specific guiatrs, pickups, etc.) a higher idling current PA, especially in a cathode biassed amps like a 5E3, will sound fat and greasy compared to low idle currents where the amp sounds too polite adn there is the possibilty of odd harmonic crossover distortion.

[MWJB]

Hi Bruce,

I agree with virtually everything you say here except...

1. Converting an input to grid leak, as you say, can make the amp sound fluffier and, to my ear, rounder. However, it works best with crystal and ceramic mics, sounds a bit gutless with CM/CR & dynamic mics. Of course, with a 5E3 type circuit you can remove the 500pf cap from the tone pot, thus making both channels the same ("Normal/Mic"), and then have one cathode biased preamp and a grid leak biased preamp. In your scenario are you keeping the cathode resistor or grounding the grid-leak biased triode's cathode?

10. Indeed, higher currents will sound greasier, but it's worth investigating slightly lower currents too. Higher currents generally also mean more feedback (especially with a 12" speaker) and more mids/highs. A slightly drier tone may be preferable for some, as well as making the amp a little easier to live with? In a couple of single ended amps running around 400v (plate to cathode) I've consistently ended up with 1K cathode resistors after listening tests, running a 6V6 at around 30mA. Equally, in a single ended EL84 amp, the more current I could throw at it the better it sounded...so all I'm saying is try it both ways.

[smokingun]

I am very interested in this thread. I just ordered a 5E3 kit from Bruce with the harp mods. I have two harp players that want this amp and we will send it through the paces to get voiced like they want. I am going to follow the mods that Bruce has suggested along with his choice of speaker. I update as the build progresses. I would like to post some clips, but I'm not sure how to do a recording. Any suggestions would be welcomed.

Greg

[Chris / CMW amps]

1) replace the 1M on the 1st input jack with a 5M resistor

Maybe OK,

BUT, if you put a .05uF cap in front of the grid load resistor, and then get the grid resistor up real high +4M7, that preamp will try and run in what is called contact bias... that does sound different and a little fluffier to me.

2) pick a channel and replace that channel's 100k plate resistor with a 470k

I'd say no

3) replace the 1k5 cathode resistor with a 2k7 (the 25uF/25V cap stays in place)

I'd say no

4) decouple that stage with a 0.047uF cap

I'd say no

5) decrease the slope resistor from 100k to 56k

There is none

6) replace the phase inverter tube with a 12ax7

The PI already uses one of the 12AX7 triodes for push pull adn that section has ZERO gain.

7) if there's a tube rectifier, use a 5u4

I'd say no

8) replace 6L6's with 6v6's

The amp already uses 6V6s.

9) install a dropping resistor to lower the screen voltage: 500 ohm, 10W

The voltage node that feeds the screen's their voltage , with respect to the plate supply is already 5000 ohms and only 2 watts or so!! Why would you need a 10 watter even if you did use 500 ohms.... somebody not using any math.
Seperate screen resistors of 2K7@2W or even higher resistance will give the PA a little more compressed sound. Maybe OK

10) bias at 27mA/tube

I'd say no...
and why pick an odd number like 27ma? That has nothing to do with what the tube is doing unless you know the tube and the B+ voltage, so don't follow any rules regarding ma of idle current without checking the other parameters.


Why to I say no to:
2,3,4,7 and 10

2., The only reason players do this is because many tweakers do not understand what the plate load resistor does.
If there is enough plate voltage delivered to the tube, a larger one can actually increase the gain.
Loosely put, gain is a function of absolute voltage drop across the impedance of the tube, and the ratio of the cathode resistance to the plate resistance, but using a way larger plate load resistor does lower the B+ delivered to the preamp tube which can make it less gainy because there isn't any B+ to work with but it will clip poorly, faster and with less tone too...
that is the wrong way to do it if you are looking for less gain.
A dead stock 5E3 has only about 105v-125v on the plate in the first place so lowering it below 85v-90v is really just silly.

3., That will bias the tube colder and can make the preamp slightly harsher sounding with more odd harmonics then if the tube is biased right in the middle of it's class A space with higher current.... yes it will have a wee bit less gain because of the lack of B+, but there other ways to get that without this method which changes the warm tone of a free running first preamp.

4., Lower impedances like to have larger coupling caps to fully pass the audio at lower frequencies so this is ass backwards.

7., Does not apply to a 5E3 with 6V6s as the amp is a lower voltage B+ amp and the 5U4 will be so lightly loaded with this topology that the B+ will be much higher... ass backwards again.

10., (Except to tweak for personal taste or specific guiatrs, pickups, etc.) a higher idling current PA, especially in a cathode biassed amps like a 5E3, will sound fat and greasy compared to low idle currents where the amp sounds too polite adn there is the possibilty of odd harmonic crossover distortion.

Bruce

This mod is mentioned on a German website (url: don't know) and maybe from Dr.Z. (Mike Zaite) AND....
for blackface or silverface Fenders!!!
I've used most tweaks AND some others (for better voicing and way less feedback) several times with great results.

[Bruce / Mission Amps]

I guess I should be spending more time with a spelling checker! ha ha ha...
Howver,

Regarding #1 and #10,
That's why I should be prefacing everything by saying "except to tweak to taste".
And I forgot to mention that the mic elements makes as much difference as the kind of pickups a guitar player uses.
With respect to the contact or grid leak bias, I assumed a builder/tweaker would ground the cathodes.
I know quite a few guitar players who like contact bias in their old amps and they don't have any special pickups.
There weren't many hot guitar pickups in the early to mid 50s and those amp guitar combos worked OK anyhow but I suspect it would sound better/different with a mic element that suits it.
By the way,... the input jacks on the old Fender tweed amps did not always say bright and normal, they frequently said something like:
instrument and MIC!
One thing is that for my pro harp player customers, no matter what I do they always want it to sound like this but ten times louder! ha ha.

[SlidePicker]

Hey Bruce, instead of using split plate loads to drop gain, have you tried using lower than usual plate loads? Or lower than usual grid resistors in the following stages?

The reason that plate loads for triodes are normally ~2x the tube's plate resistance or higher is that lower plate loads make a lot of second harmonic distortion. That's not necessarily a bad thing, though...

You'd have to closely watch the B+ voltage and plate current.

[Bruce / Mission Amps]

I've tried all kinds of things too. :eek: ha ha

One of my favorite harp tones came from a 25 watt amp I built for Kenny Dore (Maria Muldaur's group) where I used a 12AU7 for the first preamp tube with 27K plate load resistors biased solid class A but at an AMAZINGLY low plate voltage of .... 30-35vdc!
This then drove the other half of the 12AU7 at higher plate voltage, into a loose version of a Baxandal tone stack, recovered by a normally voltaged 12AY7, which then drove the other half of the 12AY7 as driver.... breathe... and then onto a pair of cathode biased... EL34s... yup EL34s.

[clyde1]

Howz about a pentode front end, like the 5687 in the Gibson Maestro, which is
a smoking harp amp? Ever used any of these Bruce? Also the 6J7 from a Masco PA. I play with an amazing harp player and he likes this stuff. He uses Astatic and Shure mics, the old ones.

[MWJB]

No real reason not to do it but if it is a 2 channel design, use a 9 pin preamp as well, as a back up. Some mics & players may have differing preamp gain needs and the flexibility of having plug in/plug out interchangeability of 12AX/12AY/12AU can be a big benefit.

[steel1953]

I've got 2, and I'd like to set one up for harp.

[MWJB]

1st stop for a Brown Concert is to make sure that your power supply is up to scratch, ensure that both main and screen filter supplies are rated for 700v minimum (B+ is usually just a few volts off 500v, which is good for power & low end). I usually use 2x100uf/350v in series for the mains (each cap bypassed with a 220K 2W resistor) and the same for the screen supply (though pick up some 47uf/350v or 450v too as 100uf's can be a pain to cram in there - size of cap is critical, you won' be able to use Spragues at the screens). You can use 20uf/500v everywhere else. Whilst you're in the doghouse ensure that your preamp dropping resistor is 56K, you can go up to 82K (ideally at 3W) so long as you remember not to drown the amp in bass (about "6" tends to be plenty). I'd keep voltages at V1, 1 & 6 at 160v-ish.

Fit Sovtek 5881WXT and bias to 7.5 to 15mA.

Sub preamp and PI coupling caps out to .1uf.

Sub Normal channel cathode bypass caps to 100uf/25v, you can also try 220uf at V1, 3.

Change 1 meg Normal channel input load resistor to 5.6Meg.

And that's about as much as they need. They're great stock but have limited low end, especially if you use tenor tuned harps and a 64 chromatic. These mods (apart from the power supply, which is "non-negotiable") just fill out the low end and brown out the tone a little.

I always like 12AX7s in the preamp but play with how many - lots of harp players just use the Normal channel 12AX7 and one in the PI. I prefer a full compliment but wind up the Vibrato channel controls to tighten up the tone if necessary (Vol full up, Bass full up, Treb up until you get the cut you want) even if you get a bit more noise.

[Bruce / Mission Amps]

The brown Concert is one of my fav amps to use for a 'loud' harp amp also.
They sound great with a mix of the C10N and P10R but I still prefer a NOS 12AY7 in the preamp.

[bob p]

Have any of you seen the multitude of You Tube videos that demonstrate classic harp amps?

One fellow in Europe started the series, which features him playing harp through his collection of vintage amps. The videos are short, and are of good production quality, with rolling titles, etc. In response to his videos, other people have made their own. Its a really cool library that's worth looking into.

[bob p]

linkage:

Harmonica Amps Vol. 1 by Harpsucker (http://youtube.com/watch?v=iZ3CBI8DUOM)

His other videos (http://www.youtube.com/profile?user=Harpsucker)

[Bruce / Mission Amps]

Do a youtube search for my buddy Ronnie Shillist.
He has a new "learning blues harp" DVD out this month.

http://www.youtube.com/watch?v=bArT1imliRk

[tbryanh]

I've tried all kinds of things too. :eek: ha ha

One of my favorite harp tones came from a 25 watt amp I built for Kenny Dore (Maria Muldaur's group) where I used a 12AU7 for the first preamp tube with 27K plate load resistors biased solid class A but at an AMAZINGLY low plate voltage of .... 30-35vdc!
This then drove the other half of the 12AU7 at higher plate voltage, into a loose version of a Baxandal tone stack, recovered by a normally voltaged 12AY7, which then drove the other half of the 12AY7 as driver.... breathe... and then onto a pair of cathode biased... EL34s... yup EL34s.

I was wondering why you chose 30-35vdc for the plate voltage? What special effect did it have? Was the headroom of the first stage affected by the lower voltage?

[MWJB]

12AU7 are well capable of running low voltages with reasonable fidelity, (especially with Bruce's 27K plate resistor & with the right mic) compared to say 12AX7. For instance if you run a circuit that has a 12AX7 running at 130vdc (100K plate resistor), then you replace that tube with a 12AY7 you will get around 90v on the 12AY7 plates. Sub that 12AY7 for a 12AU7 and the 12AU7 may run at a little less than half the voltage of the 12AY7. The AU won't compress & grind up as much as the other choices as long as you have a mic with a strong enough output to stop it sounding dull/flat.

I sometimes run a 12DW7 in V1 on a 5F6A style amp to give a little more flexibility with the amp when using different mics (100pf bright cap removed), the low mu half runs at around 40-ish vdc and still has good headroom & fidelity with CM & CR mics, some hi output Hi-z dynamics...but you would need a REALLY hot crystal to make this work.

[Bruce / Mission Amps]

...

I sometimes run a 12DW7 in V1 on a 5F6A style amp to give a little more flexibility with the amp when using different mics (100pf bright cap removed), the low mu half runs at around 40-ish vdc and still has good headroom & fidelity with CM & CR mics, some hi output Hi-z dynamics...but you would need a REALLY hot crystal to make this work.I've also used the 12DW7 with a hot mic and it works very well if you also use a split load "anode resistor pair".
On the AU7 triode, a 56K to 68K on the B+ side and a lower value at the actual plate side, like 56K to as low as 27K and then couple the audio out between the two split load resistors with something big like 100nF to 220nF.
Sock a bunch of current into the AU7 triode with a smaller value cathode biasing resistor (I recently used a 470 ohm 1 watter and a and big bypass cap... 100uF to 220uF or so).

[slidincharlie (Carlo P)]

Do a youtube search for my buddy Ronnie Shillist.
He has a new "learning blues harp" DVD out this month.

Bruce,
how do I contact Ronnie Shellist to ask about his DVD's?
You can PM or email me.

Thanks

[Bruce / Mission Amps]

Bruce,
how do I contact Ronnie Shellist to ask about his DVD's?
You can PM or email me.

ThanksCarlo I think you can just get to him through here:
http://www.youtube.com/user/RonnieShellist

or,

harmonicainstruction@hotmail.com

got-harmonica@comcast.net

RonnieShellist@comcast.net



but if not, I'll give him your email address if you want.

[Marc]

Ha! Thomas is one cool guy! I sold him the Kay 503 off ebay this spring.
It was dead and ugly.
I think he did a great Dr. frankenstein thing and made it live again!

Marc

[fogtone noam]

Hi All

I am a harp player who has just started getting a little deeper 'into it' with the soldering iron.
I've been reading books on tube amps, done some upgrade mods on my stereo tube amp and would now like to experiment mod-ing my 100 watt silver-face super-six (twin) chassis -to optimize it for harp.

Specifically I would like to try changing to cathode bias and perhaps try different (lower) plate voltages. G. Weber suggests a 600 ohm resistor and a 25-100 uf cap per each output pair. He also suggests connecting the bias resistors to ground after taking them out of the normal bias circuit.

I was wondering what people here thought about the above. Can I not just cap off the bias resistors after disconnecting ? Are the above values correct? Also I would like recommendations re. resistor values for the pre-amp tube plates in order to lower voltage .

thanks in advance - Noam

[MWJB]

Hi Noam,

If converting to cathode bias I'd recommend that you use a cathode/fixed bias switch, see londonpower.com, search under Kevin's FAQ for the schemematic. There may be some situations where the harder dynamics of fixed bias are desired. In cathode bias mode try it both with and without the cathode bypass cap per pair.

Keep the resistors at the preamp plates at 100K, it's the power supply resistor that feeds these that you want to change, maybe try 56K- 100K at 3W. I'd shoot for around 160vdc at the 12AX7 preamp tube plates (V1 or V2).

Increase cathode bypass caps in the preamp to 100uf, increase coupling caps at preamp & PI to 0.1uf. Use a 5.6meg resistor at the input jack instead of 1Meg. Experiment with bypassing the 47ohm or 100ohm NFB load resistor with a 1uf cap (gives "presence").

Your amp should be serviced (new filter caps, bias supply caps & cathode bypass caps to ensure proper operation).

Bias - in fixed bias try 15-20mA, I like Sovtek 5881 in this kind of amp. In cathode bias you may want to see around 50mA per tube, 600ohms sounds about right.

[MWJB]

Further on cathode resistor values...

Exact value will depend on your plate voltages, there was quite a margin for plate voltages accross SF amps - 600ohms would be OK for B+ over about 480v, if voltages are nearer 400-430v try 330ohms, 470ohms for voltages betwen 430-479? Plate voltages suggested are measured to ground, without subtracting cathode voltage.

If you get a little over 50mA I wouldn't be too concerned, you'll lose maybe 40-45v accross the cathode, so even if you had say 510v at the plates, 510-45 = 465v, 465v*.050 = 23W dissipation.

[fogtone noam]

Thanks MWJB, The details are much appreciated. I will take a look at Kevin's faq for making this option switchable.