FDP Home Page / FDP Forum / FAQ's

The FDP is made possible by the following companies and individual members like you.
Please use the links below to show them we value their sponsorship.

WD Music

Jensen Loudspeakers

Apex Tube Matching

Antique Electronics Supply

Yellowjackets Tube Converters

MOD KITS DIY

Sweetwater

Amplified Parts


* God bless America and our men and women in uniform *

* Illegitimi non carborundum! *

If you benefit and learn from the FDP and enjoy our site, please help support us and become a Contributing Member or make a Donation today! The FDP counts on YOU to help keep the site going with an annual contribution. It's quick and easy with PayPal. Please do it TODAY!

Chris Greene, Host & Founder

LOST YOUR PASSWORD?

......................................................................

   
FDP Jam
Calendar
Find musicians
in your area!
  Search the Forums  

FDP Forum / Amp Mods, Repairs, and Projects / More 5E8A (Low Power Tweed Twin) build question - squealing

vanceen25
Contributing Member
***

United States

Vance Erickson
Jul 31st, 2019 03:03 PM   Edit   Profile  

I got the 5E8A head built and working properly, but I have another problem / question.

For reference, it's a Ceriatone kit, which follows the original 5E8A schematic pretty closely as far as I can tell. I've made a few mods which I'll describe below (since this is going to be kind of long).

I'm hearing an audible high pitched (squealing) at some settings. It's most sensitive to high volume settings, and high treble settings, but high presence also contributes. I can get the squeal to go away by combinations of the knobs, i.e. if the treble is low enough the volume can be a bit higher and vice versa.

Interestingly, it's also sensitive to the setting on the THD Hotplate attenuator I'm using (more attenuation = less squealing), and it's sensitive to the half power switch that's installed (half power = less squealing). It's also reduced when the guitar volume is reduced.

It doesn't occur if no cable is plugged in, or if the cable is plugged into a guitar with the pickup volumes turned down.

It's definitely NOT feedback through my guitar pickups. I know that sound well, and it's different (I've tried three or four different guitars, single coils (Fender and P90's) and humbuckers, all give the same result.

Most interestingly, the squeal gets VERY much worse when a cable is plugged in with no guitar. The choice of cable matters. All guitar cables (with no guitar) give rise to squealing, although some do it significantly less than others. I tried a speaker cable, just out of interest, and there was no squeal, although there was a lot of buzz.

I've replaced every preamp tube and the power tubes, with no change at all. Putting 12AY7's in the preamp positions also doesn't help.

My first thought is that it has something to do with the lack of any grid stopper on the preamp inputs, which is part of the 5E8A design. Any other thoughts would be most welcome.

(The mods I've made: I disconnected three of the inputs from the preamp tubes and made it a single input amp. To take the unused half-triodes out of the picture, so they wouldn't make noise, I removed their 100k load resistors and their 100k mixing resistors leading to the volume controls (if I left in the mixing resistors I still had 120V on the wires to pin 7 on V1 and V2). Also, I put an on-off-on switch in one of the spare input jack holes to switch between normal negative feedback, a bigger resistor in the loop, and no negative feedback. Finally, I put in a tone stack disconnect switch in one of the other spare input holes, to disconnect the treble and bass grounds for gain and mid-boost.)

vanceen25
Contributing Member
***

United States

Vance Erickson
Jul 31st, 2019 09:22 PM   Edit   Profile  

I should have mentioned that I used shielded cable from the input jack to V1 and from the treble pot to V4 pin 2 (makeup gain stage). The jumper wire between pins 2 on V1 and V2 is shielded, but not grounded.

pdf64

UK

Aug 1st, 2019 02:25 AM   Edit   Profile  

I suspect that your mods are causing the problems, perhaps due to layout and bias shortcomings.
Was it ok before you did the mods?

5e8a Schematic

(This message was last edited by pdf64 at 07:49 AM, Aug 2nd, 2019)

Peegoo
Contributing Member
**********
**********
**********
****

If irritation occurs

discontinue use.
Aug 1st, 2019 06:54 AM   Edit   Profile  

pdf64 makes a great point: gain circuits have to be carefully designed to prevent oscillation.

If you built the amp from the ground up incorporating the mods, you don't know which change may be causing the oscillation.

Leftee
Contributing Member
**********
**********
**********
**********

VA

The Escalator
Aug 1st, 2019 08:10 AM   Edit   Profile  

FWIW - if I use coax in an amp circuit I always ground the shield. I don't know if this will help your issue, but it's a good practice.

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 1st, 2019 08:15 AM   Edit   Profile  

Peegoo, pdf64, thanks for your responses.

To be honest, I don't really know if it was OK before the mods. I had it in my house and it wasn't convenient to turn it up really loud.

The tone bypass mod just runs the bass pot ground wire to a switch at an unused jack hole, and then to ground, either directly or through a 620k resistor. The ground normally passes along this path anyway.

The NFB mod does involve running a wire from the impedance selector to the front panel switch, a pretty long stretch of AC signal. I could use a shielded wire for that if it turns out that disconnecting helps.

I could easily disconnect the jumper between pins 2 of V1 and V2 at V1 to see if that helps. If so, I could arrange a ground for that shielded wire. I'm thinking it ought to connect to the insulation on the wire coming from the input jack to V1 pin 2, to avoid ground loops. That might be a little tricky, but it's do-able.

One other thing occurred to me. It seems strange that reducing the volume with the attenuator reduces the squeal, since you wouldn't expect that to affect any internal oscillation. I wonder if there some RF weirdness on the output side of the attenuator that's feeding back through the cable and/or internal wiring.

I'm waiting on some parts, and I'll check out these ideas in a few days. I'll report back here in case it helps someone else. If either of you have any more thoughts, it would be most welcome.

pdf64

UK

Aug 1st, 2019 02:26 PM   Edit   Profile  

I suggest that you don't use the spare input jack holes for anything that isn't connected to the input stage/s.

Putting circuits that have lots of gain between them into close proximity will tend to cause problems, breaks the 1st law of good practice in regard of layout.

The issues are probably worse with a reactive load (speaker) than a resistive load (Hotplate set to attenuate) because the amp gain will tend to track speaker impedance, which being inductive, steadily increases above nominal above a few hundred Hz, at the rate of +6dB/octave.

(This message was last edited by pdf64 at 09:02 AM, Aug 2nd, 2019)

willie

Too Near Atlanta GA

Amp Tech Emeritus
Aug 1st, 2019 02:33 PM   Edit   Profile  

Agree totally with comments made...I would reverse engineer that amp back to original spec and see what gives then...

w

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 1st, 2019 02:44 PM   Edit   Profile  

Thanks all, especially pdf64 for the comments on layout. The negative feedback could be too high gain for that area. It should be easy to check; just move the switch. I wouldn't wouldn't expect that the ground from the treble pot would have much gain, but I could of course be wrong.

Watch this space...

pdf64

UK

Aug 2nd, 2019 05:49 AM   Edit   Profile  

NFB related controls, eg presence, are usually at the other end of the control panel if part of the original design; if mods, they are generally best placed close to the speaker jacks, away from any preamp circuits, especially early stage.

Regarding the tone stack lift switch, when it's in the 'normal' mode, then yes, probably no problem.
But when in the 'lift' mode, assuming 12AX7 gain stages, there's 2 stages each of a gain of up to ~60 in cascade, so total of ~3600, ~70dB (ie with controls maxxed)! And to make things even worse, the signals are in the same polarity.
So 1mV at the input causes 3.6V in a circuit less than 1" away. Given the circuits are high impedance, only a little capacitive coupling is required for that to cause another 1mV to appear at the input, and hey presto, you’ve built an oscillator!
There will certainly be some degree of positive feedback going on there; it's uncertain whether it would actually cause free running oscillation, but it's going to have some tonal effect and is definitely not compliant with good layout practice.

Have you considered the biasing of the input stage? ie the original design has the 4 input stage triodes sharing a 470 ohm cathode bias resistor; my understanding is that you now only have 1 triode?

(This message was last edited by pdf64 at 09:13 AM, Aug 2nd, 2019)

Peegoo
Contributing Member
**********
**********
**********
****

If irritation occurs

discontinue use.
Aug 2nd, 2019 06:32 AM   Edit   Profile  

"If I use coax in an amp circuit I always ground the shield."

Only one end of the shield. Not both :o)

pdf64

UK

Aug 2nd, 2019 07:01 AM   Edit   Profile  

A shield that’s not connected to anything is pointless, no different to a regular piece of wire.
To avoid hum due to HT 0V ‘ground’ loops, the shield should only be connected to 0V at one end if the circuits at both ends are already connected to 0V.
If one end doesn’t have an 0V connection, eg insulated input jack, then both ends of the shield must be used.

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 2nd, 2019 08:35 AM   Edit   Profile  

Thanks, pdf64 and Peegoo.

I have the squealing whether the tone lift is engaged or not. It does get worse whent the lift is engaged. I've got a Bassman AB165 head in which I've wired the tone lift to the "Deep" switch on the Bass channel, and I've never had any issues there. But I know this could be different.

I have two triodes running, V1a and V1b. At first I left the cathade resistor the same, since four triodes would be conducting, even if two had no input. In order to reduce hum, I disconnected the B+ from V1b and V2b (by removing the 100k load resistors and mixing resistors). After that I changed out the 470R cathode resistor for an 820R resistor. I know that to keep things the same it ought to be 940R, but lots of amps use 820 (or 1.5k for two tubes).

All the shielded wires except the jumper between V1 and V2 pins 2 are grounded at one end (not both). I knew that to do it properly I should connect the jumper ground to the shielding for the input lead to V1 pin 2 (so that both would be grounded at the input, avoiding ground loops). Frankly, that was going to be a hassle. If I track down the squealing to the jumper, I'll get it grounded.

pdf64

UK

Aug 2nd, 2019 11:19 AM   Edit   Profile  

The tone stack and deep switch of the AB165 are in the plate circuit of the input stage.
So 'only' a gain of up to 60 (35dB) and signals in opposing polarity between a tone stack lift switch mounted in the #2 input jack hole, and the #1 input jack. You'd have to try really hard to get oscillation in such a set up :-)

Plus I think that the input jack holes may be mounted further apart in a BF than a Tweed chassis? And tweed chassis are pretty cramped, usually lots more space available in a BF, so layout in a tweed becomes far more critical, all else being equal.

So a rather different scenario with your AB165 than the 5E8a.

(This message was last edited by pdf64 at 02:33 PM, Aug 2nd, 2019)

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 2nd, 2019 02:14 PM   Edit   Profile  

OK, got it sorted.

Before taking everything to stock, which was good advice from several of you, I tried a few things.

First I grounded one end of the shielded jumper from V1a to V2a. It meant "grafting" a short ground wire into the outer braid of the wire coming from the input to V1a, all heat-shrinked up. Full volume, full treble, no squealing. But diming the Presence caused squealing.

So I replaced the wire from the impedance selector to the switch with shielded wire, grounding it at the input jack. Full Presence, no squeal.

Thanks again to all of you for your help!

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 2nd, 2019 02:20 PM   Edit   Profile  

pdf64,

Just saw your additional comment about the AB165 vs the 5E8a. Obviously you're right, and the point about the signals being in phase after going through two triodes in my setup (unlike the AB165) is something I didn't think of.

Maybe I've gotten away with it by using a shielded input wire. I'll watch out for any signs of instability.

Thanks agains.

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 5th, 2019 08:39 AM   Edit   Profile  

Looking back over this thread, I realize I left out some relevant facts and made at least one mis-statement. I want to correct that so this might be helpful to someone with a similar problem.

First, when I listed the mods I had made, I neglected to mention that I had added a jumper from pin 2 of V1 to pin 2 of V2. This makes it possible to use both volume controls, just like running a jumper across the input jacks in a four-holer.

Second, at one point I said that I was running triodes V1a and V1b. That was wrong. It was V2a and V2b.

I hope this helps anyone who might have been trying to figure out what was going on.

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 6th, 2019 06:33 PM   Edit   Profile  

Sigh... I meant V1a and V2a. I've probably confused everyone enough.

pdf64

UK

Aug 12th, 2019 08:46 AM   Edit   Profile  

Thanks for the clarifications, great that it's worked out for you.
You may wish to consider screening around the jack / switches; one method is to make and fit 'L' plates inside the chassis, to isolation one sensitive circuit from its neighbour. Easy to do with very thin sheet metal.
Kinda like Vox did with the 60s AC50, to keep input jacks and speaker output terminal block from interacting.
See http://www.voxac50.org.uk/images/AC50s/large/early_25/closeups/early_25_06.jpg
http://www.voxshowroom.com/uk/amp/ac50tr_hood.html

vanceen25
Contributing Member
***

United States

Vance Erickson
Aug 13th, 2019 09:55 AM   Edit   Profile  

Thanks very much, pdf64. I may not need that on this amp, but I'm considering modding my Bassman AB165 to use that spare triode, which will involve a lot of gain. Screening will probably be handy there.

FDP Forum / Amp Mods, Repairs, and Projects / More 5E8A (Low Power Tweed Twin) build question - squealing




Reply to this Topic
Display my email address             Lost your password?
Your Message:
Link Address (URL):
Link Title:




Moderators: Chris Greene  Iron Man  reverendrob  

FDP, LLC Privacy Policy: Your real name, username, and email
are held in confidence and not disclosed to any third parties, sold, or
used for anything other than FDP Forum registration unless you specifically authorize disclosure.

Furtkamp.com 
Internet Application Development

Copyright © 1999-2019 Fender Discussion Page, LLC   All Rights Reserved