Hiwatt problems

[frank57]

I'm having hum and buzz problems with a korean Hiwatt custom 20. You can read about it here
http://www.electro-tech-online.com/repairing-consumer-electronics/99783-hiwatt-custom-20-tube-amp-hum-problem-2.html#post825711
Right now I'm re-twisting the wires and waiting for parts,
but I wanted to replace the jack which I was told was not a shorting jack.
I tried a marshall and fender jack but they are too small.
Is it okay to try a cliff solder lug jack and cut it to fit?
The s2-1220a or s4-1322?
They used a very strange looks like custom reverse l-leg jacks.
They are about 9mm long.
I could try swapping another jack that's in there as well.

[PRR]

> I was told was not a shorting jack.

That only matters when nothing is plugged in.

If you have two jacks, shove a pot-shaft in the un-used hole, that WILL short it for testing.

> cut it to fit?

If you cut the end the plug goes into, the plug will go too far, past the contacts.

[PRR]

PCB image

That PCB layout is very poor. It is bound to have residual buzz.

How much do you LOVE this amplifier? If "not much", I suggest you sell it to someone else.

The layout might be fixable. I'd have to see the whole chassis, to know if stuff would fit. If I did the thinking, cutting, and re-connection, I'd want to charge nearly what asian-made EL84 amps should cost, so I doubt it is worth doing.

[frank57]

So it's not a good idea to try replacing the jack?
Where is the buzz coming from exactly?

I think they had low pass filters in here on the input and output of the preamp which did a good job of filtering it out. They were removed and I got buzz. I was told it's high end static.
Would a 68k grid stopper on tube 1 help?
I wanted to replace the caps and see if that would help with the hum.
I could post some shots of the chassis.
I do like the amp and it is a hiwatt style circuit, but if it's a lost cause or too costly I was looking at Jet city amps.They have a very similar sound.

How tough would it be to make a handwired version of this Hiwatt?
I could use the two transformers and the chassis . It's 16.5 inches wide.
Probably have to punch some holes I think. But I would need a layout.

[imaradiostar]

Are you in love with the sound of a hiwatt?

Why not build something simpler in the chassis? EL84's don't need much signal and can sound great in a circuit with only 2 12ax7's while retaining all the same front panel controls your amp has.

That said, you could build a hand-wired hiwatt circuit in the chassis with a little effort.

jamie

[frank57]

I do like the Hiwatt sound and this amp does have some of it but with more gain.
Something simpler might be a good idea.
Right now I'm going to try new filter caps for the hum, if I can get the ones in there out.
They glued them flush with the board and are proving to be a major headache.
So far two out, three to go.

If that doesn't work not sure what to do with the amp,
If there was a handwired layout for this amp I would try it.

They put the tubes and transformers tight in a row so probably would have
to punch a new hole or two and move a transformer back.
I was looking at the ceriatone boards but they are el-34 and a little too powerful .
Or the triwatt from trinity amps. They do make the chassis complete but they're too big to fit in this amp.

Otherwise I was thinking of the new soldano jet city amp.

[daven]

I do like the Hiwatt sound and this amp does have some of it but with more gain.
Something simpler might be a good idea.

old AX84 project:

http://www.ax84.com/index.php/oldprojects.html?project_id=lowatt

you'd have to pick one channel (or switch components with a DPDT switch), and perhaps ditch or hardwire the presence control since you don't seem to have enough knob space.

btw, I built a SE EL34 version into dead SS crate amp for a friend.  when he plays it, it's like 70s arena rock heaven (but not as deafening).

[PRR]

> Where is the buzz coming from exactly?

Any place it can.

And the PCB has such obvious "poor design" that I wonder if there may be other problems.

What hits me: if Standby Sw Blue is the PT HV CT, then the raw spike-current flows through about 3 inches of PCB trace before it hits the main filter cap. The trace-length has significant ripple voltage induced in it.

The preamp filter caps are to the left of this trace. The signal ground comes off the right end, goes around the board to the input.

While looking around: R23 probably shouldn't be 220K, though this would not buzz.

[frank57]

That Lowatt is really interesting!
I don't know if I can pull it off without step by step instruction.
It's a more precise version of what's in there now I think.

Could a handwired layout be done for this Hiwatt Custom 20 or a new pcb layout?

Here is the schematic that Niles Goodwin drew out from my Board drawing and I touched up.
I think the idea is:
 V1 is the Brilliant channel or input(from DR504 OR SA412).
V2b is more or less the Normal channel moved over.(maybe 2k2 might be better for r5?)
The tone stack is a little different, maybe we can make it more similar.
At some point someone disconnected two traces to take out v1 in an attempt to reduce the hum.
Doesn't work. You reduce the gain, but does nothing or very little for buzz and hum.
They are reconnected.

[frank57]

Here's an old shot of the board from the other side so you can see the bias switch.
A bias pot has been added and c23 was changed to 100volts.
I've retwisted the wires a lot better than what you see here, but haven't resoldered them back in just yet.

There are some "repaired" connectors on tube 2 that might be part of the problem.
I will post a shot later today.
Also the line out connector has broken off.
Could I just solder the wires directly to the pcb?
There would still be connectors on the other side though.

The mains switch receptacle seems a little loose too.

[frank57]

Here are the "repaired" connectors.
Connectorsweb is the ground for tube 1  the other pic  is for tube 2.
One of them I had to repair myself because it was barely hanging on.
Maybe source of hum or buzz ?
I can't seem to find these connectors anywhere.
Maybe Mouser might have something.

So far I have one cap left to take out.
They did a crap job putting them in making it very tough to take them out.
They gouged the board somewhat where these snap in caps are and glued the crap out of them.
Also some resistors and caps have the leads barely poking through the pads.

[frank57]

Can uf5408 or uf4007 diode be substituted for IN4004?
If like I saw on one site, the buzz might be coming from the diodes, this might help, I think.
I read somewhere faster diodes can help.

[PRR]

Move the blue standby wire to the negative lug of the main filter cap C22 100uFd 450V. I mean TO that lug, even though this will mean a blobby butt-solder joint.

There are many other PCB problems, but when I started trying to work it out I got a headache. Combined with general low-cost "quality", I'd be real tempted to bulldoze the chassis and start over with proper turrets and sockets.

Getting the massive ripple-garbage localized within 1/8" of ground bus, instead of spread-out over two inches in the middle, may help a lot.

If you do not understand exactly what to do, you can smoke the whole amp.

[imaradiostar]

Can uf5408 or uf4007 diode be substituted for IN4004?
If like I saw on one site, the buzz might be coming from the diodes, this might help, I think.
I read somewhere faster diodes can help.

It's very likely that the diodes aren't the problem.

Read most of this pdf document:

http://www.freewebs.com/valvewizard2/Grounding.pdf

The PC board on the hiwatt design is what results when accountants and computers create a layout for a design that might otherwise be a great circuit. Properly designed a PCB can be fine. That one has quite a few questionable things- if it were my friend's amp on my bench I'd have ripped it out already!

I agree with PRR (as we usually should, he's smart). If it were my amp to fix I'd remove the 1n4004 diodes and resolder them (or new diodes) so that they'd be soldered at the + lug of the 100uf cap. Same goes for the center tap of the output TF. I'd also solder the ground lead from the power TF (dunno if this is from the standby switch or a direct connection) directly to the negative lug of that same 100uf cap.

So basically what PRR said... 

Many users here are happy to help make layouts for a design and share it so we can all benefit. If you're not up to the task of making a board you might be able to convince someone here on the forum to make one for you. Working on tube amps can be very dangerous so if you're not comfortable don't get in over your head.

jamie

[frank57]

So putting in uf4004 wouldn't harm anything anyway?
I remember one tech seemed to say he saw buzz coming from there.
But these guys also said "it's fixed".
Would a 68k grid stopper help on tube1 for hum?

I can easily move the standby switch wire and solder it.
This is right up my alley.
I don't know if it's long enough to reach though.
So I just solder it underneath the board to the negative lug of the 100uf cap?
The caps were snap in, but they seemed to have gouged the board a little.

But just to make sure:
The standby switch was added on it was not original.
The black wire from the power transformer was cut and connected to the top lug of the switch.
 The blue wire goes to the board from the bottom lug.
 Maybe you can't see it clearly in the photo?
I can shoot another one.
All the big caps are off the board right now.
Since a bias pot was added the switch tends to make a bit of a powering down type sound for a second.

The green wire from the power transformer goes straight to the ground lug on the right in the photo.
Also green wire from mains receptacle on same lug.
The diodes might be too complicated for me, but I could move the transformer wires.
The OT is on the left and has a red wire going to the board.
 
Could the hum be also from the fact it's all ac powered as well?
There are no resistors at all on the tube pcb.

I could try these things out on the amp and start a separate chassis on the side .
If someone could do a clear layout of this amp I would try it.
Or if someone might be interested to build it.
I think I could handle the preamp board.
I could always get  voltages and so on checked by someone else.

[frank57]

Here's the Hiwatt sa412.
Korean Hiwatt sort of based on this amp?
Although quite a few Hiwatts are similar in their general approach.
But with the normal channel on tube 2b.
Brilliant channel tube 1.
At least that's my guess.

[PRR]

> Could the hum be also from the fact it's all ac powered as well?

Well, yeah. If you went out in a field (far away from all power) with 400V and 6V batteries, there would be no hum.

Hum/buzz is all about being too cheap to buy $100 of battery per hour. Oh, and liking to have light, heat, etc.

The primary problem _I_ see is the PT CT lead (black) getting to the main filter cap by the long path through PCB trace. The PCB is not a perfect conductor. There is significant ripple from the black/blue lead to the main filter cap. You have preamp power filtering to the left, preamp signal path to the right and around. So there is significant ripple injected into the preamp (all the way through the driver).

I don't mind where the rectifiers land. The layout to the OT CT is fine. I suppose R37 gets a little more ripple than necessary, but its job (with C20, which is in the line-of-fire between PT CT and C22) is to reduce ripple. That part's fine.

The 68K has nothing to do with hum/buzz. It stops radio leaking in. It's a good idea. It's not your immediate problem.

[frank57]

Is it a good thing they did putting the green wire from the Power transformer
 and the green wire from the power cord
receptacle on the same bolt?(*The green wire from the Pt is strictly a ground wire?(I know dumb question))

There was something I saw about twisting the center tap wire around the
red high voltage wires that might help reduce hum as well. Something to do with the pulses.

If I put in a 68k resistor afterwards, where exactly would I put it?
Would changing r9 and r14 to 1watt resistors make some sense?
They are 1/2 watt right now. R37 is actually a 5watt resistor as well not 3watt like on the board.

Putting in uf4004 diodes good or bad idea?
Makes no difference?

There was one guy on a site I saw who suggested running the jumper wire on the trace side on the input,
straight to the ground bolt where the chassis is grounded.
Does that make any sense?

[frank57]

Shot of where bias pot was installed on board.
This tech wrote all over the board for some reason.
One thing he mentioned was too much voltage on tube 3.
Mean anything?
Does it matter if replacement caps are a bit smaller (width or height)?
I see some by Cornell dublier at Mouser that might be good for the  100uf.

[frank57]

Quick question:
22awg 600volt wire was used for the black wire going from the transformer to the standby switch.
The blue wire going from the switch to the board(will be going to negative lug of the main filter cap C22 100uFd 450V) is 18awg. Is it ok to use the blue wire as is or should I replace it ?

[frank57]

Here's a possible solution from Paul Ruby of Paul Ruby amplifiers.
It's very much like PRR's suggestion but goes further:

I can't say this is the problem, but I'll take a stab...

Solid state rectification causes very large ground current spikes every 8ms
to charge the first filter cap. If ANY of this current is in the same ground
wires as signal, it will get into the signal path and be amplified. With the
main rectifiers and the first filer cap right in the middle of the amp, it's
ground path is right in the middle of the main ground buss of the amp (the
horizontal ground track running the length of the PCB). And, the HV mains
and the CT fly right over and near the preamp tubes and you can get magnetic
coupling as well. This is TERRIBLE design.

The HV wires (red), recto diodes, first filer cap and stand-by switch MUST
be well isolated from the rest of the amp. Here's a representative picture
of one of my amps with solid state recto. All these components are on a
separate board mounted directly to the power transformer bolts for shortest
possible wire lengths and physical separation from everything in the signal
path.
http://paulamps.com/MartaGuts2.jpg

I would not add grid stoppers as those will make each grid more susceptible
to picking up the buzz transmitted by the HV recto circuitry (it's literally
an AM transmitter loop of 120hz buzz). I know this will be major surgery but
I would strongly suggest moving the recto diodes and first filter cap way
off the right side of the image (img_0514.jpg). Bolt some terminal strips to
the two bolts over at the far right side and do all the HV rectification and
first filter cap right there and then jumper over to the main board,
including the ground line. The PT HV lines and CT all go directly over to
the recto and first filter cap FIRST in a nice tight twist running along the
corner of the chassis, away from the tubes and signal path.

Another problem with the amp is that the input jack is not directly grounded
to the chassis. I always make my single, STRONG connection to the chassis
right at the input jack. This keeps 120hz buzz from coming into the amp from
outside via the shield of your guitar chord. The layout I see in your
picture shows that this buzz will flow directly along the guitar shield and
through the shield of the line jumping to the tube and into the 12ax7. The
irony of this is that a grid stopper on the first tube will stop this from
getting into the tube BUT will allow the internal recto buzz to be
transmitted into the first tube grid. This amp layout makes you directly
susceptible to one or the other. Find a way to get a very short, strong,
direct connection from the input jack ground to the chassis an remove the
existing, stupid long wire doing this now. Here's a representative picture
from one of my amps that does this. It's a bit hard to see but I have a lug
bolted right to the aluminum side bar adjacent to the input jack. That is my
single connection from circuit ground to chassis ground. It blocks incoming
RF on the guitar chord and allows me to NOT use a grid stopper on the first
tube so also allows immunity to RF pick-up right at the tube grid.

http://paulamps.com/TimSmithGuts2.jpg

I think if you do the above stuff, you have a good chance at solving the hum
and buzz. Both are big problems and you're just peeing upstream trying other
things first. There still maybe other things to do after this, but these
come first.

- Paul

Okay! So how do I go about doing it?
I need some kind of layout or guide.
I'm a little uncertain how to get back to the pcb and how to do the terminal strip.
All the caps are off the board and wires desoldered, so it's ready to go!
The transformer bolts are 3 inches apart, bolt to bolt.
2 and 7/8 "centers to center.
Might have to flip the standby switch so that the wires run to the right hand side too.

Also what can we do about that input jack?
I could always get someone to do the terminal strip but it depends the cost.

[frank57]

Anybody know how to implement paul's idea?

[tubesornothing]

I think your not getting many replies because to rework a PCB amp is a pain, a total pain.

As PRR says:

There are many other PCB problems, but when I started trying to work it out I got a headache. Combined with general low-cost "quality", I'd be real tempted to bulldoze the chassis and start over with proper turrets and sockets.

[frank57]

I know. I know.
But I can't get a layout to handwire it.
Maybe the Lowatt design could be modified but I can't follow the drawing though.
It's not clear enough for me. The tube pcb would probably have to be scapped as well and go straight to the tubes.
The chassis is small so I don't know if it could be handwired.
It's 16 1/2 by 8 1/4 by 2 1/4.
 The pcb is 11 3/8 x 4  5/16 inches.
One crazy idea might be to redo the pcb and go half handwired like on the hiwatt custom 50 a couple of years ago.

I don't think Paul's idea is too tough to do though.
Might be the only hope.
Everything is ready to go.
I did a rough drawing trying to visualize it, but don't know exactly what goes where.
Would putting a bolt on the chassis near the jack and running a short 10mm wire to it ground it to the chassis?

[tubesornothing]

The primary problem _I_ see is the PT CT lead (black) getting to the main filter cap by the long path through PCB trace. The PCB is not a perfect conductor. There is significant ripple from the black/blue lead to the main filter cap. You have preamp power filtering to the left, preamp signal path to the right and around. So there is significant ripple injected into the preamp (all the way through the driver).

This is most likely the problem with the amp.  If you want to ground jyour input jack, just grab a short aligator clip, clip it to the ground of the jack, and then to a bolt close to the input jack.  If the hum goes away you have solved your problem.  If it does not go away, its the next step.

Here is a bunch of info on handwired hiwatts.  http://mhuss.com/Hiwatt/tech2.html

[frank57]

I think paul and prr are both saying the same thing in a way.
I think Paul's solution is more complete by moving all the garbage and diodes off the board
and moving the wires to the corner.
Then coming back to the board.
I don't know how to do it technically though.
Where do we reconnect on the pcb?
I could try prr's simpler solution of putting the blue wire on the negative lug of the filter cap too.
They used 18awg wire for it, which I think might be better changed to 22awg.
I'd like prr's opinion on it all too.

The grounding of the input jack is more for the buzz than the hum.
I would have to put a bolt near the jack and run a wire to it.
Would that work or would it create a ground loop?

If anybody wants to do a handwired layout for this amp I could try it.
But there is a problem with size I think.

If it all fails I was looking at the jet city amps.

[tubesornothing]

You have been given lots of suggestions, its time for you to make a decision and try something.  I recommend trying the easiest solution and see where it gets you.

[frank57]

If I ground the jack to the chassis, do I then disconnect the long shielded black wire going to the tube v1?
It has a shield on one connector and the wire on the other going into the molex connectors on the board.
So I would simply pop them out.
Would I leave the jumper in place on the trace side?

[frank57]

Can this amp actually be reworked as a handwired amp or is it impossible because of the chassis?
How much would it cost?
Here's how Hiwatt used to do their pcb amps.
This is the right way. Boy if the photos were better I could almost trace it out!

[tubesornothing]

You'll have to do some leg work and find an amp model that will work for you, get the schematics and layouts and see it it will fit.  Check the mark huss link I gave you above to get started.

Once you figure out if it can fit, then you can decide if you have the time, money and skills to be able to do the job.

[frank57]

You know most of the preamp boards for these Hiwatts are 11.75 so I think that's ok.
I'm going to try to do the preamp layout for this amp and see what comes out.
The filter caps would probably have to go on the side I think.
Would a tech have difficulty doing a handwired job from the schematic of this amp?
But first things first, I'll try to get a layout of Paul's suggestion.
Or try Prr's idea.
Right now I have to wait for parts.

[danhei]

Having built a Hiwatt clone I would say read all the info on Mark Huss's site. And, from what I understand, they can be prone to noise so I would recommend following the actual layout, which is at his site, as closely as possible.

[frank57]

Well my brother took it to someone and he moved the bias pot on the outside.
Is this a good or stupid thing?
There's a lot of heat around.
How to lock that pot from moving?
It's just a regular pot.
I'm leaning towards selling the thing and moving on at this point.
What say you guys?
It's not a terrible sounding amp and there is some hiwatt in the design of the circuit.
It's noisy that's all.
There was a suggestion made to ground v1 and v2 to the back bolt and also cutting the board ground track before v2. Also suggestion to take out v1 again.
I think the poor layout simply overwhelms everything and these ideas won't do much.
V1 was taken out before with small improvement in hum.
I grounded the input jack to the back bolt and perhaps there was perhaps a small improvement in the hum.

You can eliminate the hum and buzz but you would have to bias it fairly cold.
The buzz is primarily I believe from the gain control.
I think around 30 percent of the plate limit.
With the resulting drop in volume etc.
Using an Amphead bias probe, I can see the plate voltage is high for an el 84 amp.
Around 370 to 380 range. Using an amphead probe system it should be biased between 15 to 21ma or so.
15 being 50% of the plate limit, 21 being 70.

[The_Gaz]

If I was you, it seems like I would be pretty inspired at this point to build my own, and really start trying to understand how every thing's working. In fact, that's really how I came into amp building - getting the sh*t frustrated out of me by a bunch of shoddily made tube amps. I guess I owe all this fun to poor PCB layouts, and cheap tube sockets 

[frank57]

Yes I'm looking into it.
I mean there are other things that could be tried on this amp,
but with the amp working I'm inclined to sell it.
I'm going to check it out a little further.
If i knew how I would rebuild it.
What should the right bias be?
Around 50 percent of the plate dissapation?
Not sure how to lock that pot.
I don't think a ceriatone layout would fit in here.

[HotBluePlates]

What should the right bias be?
Around 50 percent of the plate dissapation?

Assuming it's a class AB amp, and with no other information, bias to 70% of plate dissipation. You can run cooler or hotter, as long as you don't get red-plating while playing.

Not sure how to lock that pot.

You can't, because it's not a locking pot. You adjust it, and then don't touch it. It won't move without help from you. 

There are real locking pots, and they have a nut which, when tightened, activiates a clamp that keep the pot from turning. Your guy didn't use one because not a ton of places stock it, and it will cost you $11-15 depending on the vendor, instead of the $1-2 the pot used actually cost.

[frank57]

The plate voltage tends to be around 380 right now.
The buzz disappears if I remove tube 2.
The input jack is now grounded to the back bolt where the pcb is grounded.
Nothing wrong with any components.
Only problem is 50 to 70 percent dissapation I get more buzz and hum.
Did he put the pot in a good spot? Lot of heat there.
I found a locking pot here:
http://angela.com/clarostat25kohmlineartaperlockingshaftminipot381-l25k.aspx

The last guy used some kind of epoxy to lock it when it was inside the amp.
Would a dab of loctite or krazyglue or nail polish work or it will burn?
I'm just worried it will move.When I got it back from this last tech it was at 48ma in other words it was over the plate limit.

[Shrapnel]

The last guy used some kind of epoxy to lock it when it was inside the amp.
Would a dab of loctite or krazyglue or nail polish work or it will burn?
I'm just worried it will move.When I got it back from this last tech it was at 48ma in other words it was over the plate limit.

You need to ask yourself or a series of questions about this: Will I ever need to re-bias? Will I ever change power tubes (re-biasing is STRONGLY recommended)? Will such a locking method allow me to come back, and without damaging and/or replacing parts, undo the "lock" to re-bias?

On this last question apply it to each method, or substance, you're thinking of locking it with.

As far as it being off, and idling too high, this may have been the tech's fault in setting it instead of "bumping" it off. Did the tech set it, let it run/idle for at least a half hour and check and set again? (Did you allow the time?) Anything's possible, and yes locking it down would let you know if it was a bad bias job.

[frank57]

The original tech installed it inside the amp and set it with glue of some kind.
The last tech believe it or not just left it loose.
Yes it does move , I just played on the amp and accidently touched it and everything was off.
From what I hear the hum and buzz seem to all increase from v2.
Take v2 out everything is acceptable.
So grounding the input jack to the back bolt seems to have had an effect perhaps.
I'm leaning towards selling it at this point because it's just been exhausting to tell you the truth.
Maybe grounding v1 and v2 to the ground bolt might help but you really want to take out a drill on this thing?

Boy these mesa boogie russian tubes are crap! Blech is all I can say.
I usually use the new tungsols a mullard 12at7 and a rubytube in this amp.
Gives a nice warm sound. Mesa boogie Sovtek is very splotchy distortion.

[frank57]

Anyway to see if grounding v2 without drilling and cutting the earth track will work?
Looks like the ripple and buzz are getting in there.
If that's not it then it's the wiring. Nothing you can do there.

[frank57]

Here's a possible fix from Phil Abbot.
Maybe could just do v2?
The input jack is now grounded to the back bolt and the ground loop is gone.
Hum and buzz increase withh v2 in there.
How complicated is this to do and is it worth the bother?
How much would someone charge?
Not sure I can do it myself.
Loctite works fine on the bias pot.
You can easily break it by twisting.


I'm posting it in case other people have the amp.
Still will probably sell it.