Hiwatt/Ampeg Clone

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,
I am building an sort of hybrid 200W Hiwatt amplifier including the midrange circuit from Ampeg SVT. The amp has two channels with Hi/Low. You can see the circuit attached
I just assembled the whole amplifier, and for some reason I have two issues
.- There is an oscillation, like something like a chopper sound, which appears when the volume/treble/master increases. Also, when I play with the midrange rocker switch, the same chopper sound appears
.- There is a dependency between the volume at each channel. I was expecting as I introduced a valve V3 to isolate both channel.
Apart from that, the sound is very clean and there is almost no hum.
Does anyone has an idea about how to solve this?
 

Attachments

  • output.png
    output.png
    86.4 KB · Views: 488
I see that the wipers of your volume controls are connected directly to the grids of the 12AU7 which are floating above ground potential.
Suppose you run into trouble here, as the position of the volume pots have influence on the bias of the 12AU7s? Guess you would have to add decoupling capacitors.

Cheers
Georg

btw, in your schematic, the KT88s' control and suppressor grid connections are swapped. As the amp basically seems to work in reality this seems to be just a drawing error.
 
Thank for your comment. Indeed, I thought the same while I was doing the design. However, the original Hiwatt circuit connect the Master directly to the next pre-amp stage.

DR_Pre4Input_v0.gif


Actually, C20 and C21 are the ones which are decoupling from HT to the 12AU7 grid.

I disconnected the circuit of the midrange, and the chopper-like oscillation is gone. I suppose there is something wrong there. Also, for some reason the Presence does not affect the audio signal at all

About the KT88, the schematic is not correct. I need to check that, but in the real circuit all is fine.

Do you have any other suggestion?
 
Actually, C20 and C21 are the ones which are decoupling from HT to the 12AU7 grid.

Hi blackwhaleamp,

it is as I told you.
Unlike in the Hiwatt, the grid of your 12AU7 is floating above ground because auf the 13k cathode resistor. When turning the volume control down you pull it to ground. Got me?
Take a multimeter and check it by yourself. You need to decouple the grid(s).

Alternatively, you could decouple the whole tone stack including the volume control from ground, but I see no advantage in doing that.
You need to fix this mistake, chances are your filter section works as expected then.

and one more thing: The grids of your V3 have no grid resistors. (1M from grid to ground). Add them.

Cheers
Georg
 
Last edited:
Yes, looks better now regarding the input of the EQ stage. But the the schematic still lacks the grounding resistors on both grids of V3. I strongly advise you to put them there. With unterminated grids, V3 may be pushed into an undefined operating state.
Adding additional filter stages to the HV supply is certainly a good idea too, but check the plate voltages on the EQ and filter stages afterwards. If the sound suffers from too little headroom you can try to make the filter resistors smaller. Not sure if eg. R58 is necessary, you could try with and without an decide by the plate voltages you reach on the input and EQ driver tubes.

Cheers,
Georg
 
Last edited:
I rearranged the circuit as the one attached. So, the motorboating (aka put-put-put) is gone :). However, I still have some issues

1.- The amplifier is not as loud as it should. I am not sure if I am missing one additional pre-amp stage, or simple there is another thing wrong with the amp
2.- After the modifications, the pre-amp was not working. So, I connected the scope to C28 and C30 and suddenly it works :confused:. So now the midrange circuit works (with some noise)
3.- The PRESENCE is subtle, almost unnoticeable.

I will add the 1M resistors to the grid V3 tomorrow, and see the effect

Thank for the help :)
 

Attachments

  • output.png
    output.png
    61 KB · Views: 361
So, I connected the scope to C28 and C30 and suddenly it works :confused:.

This really does not come as a surprise. By connecting the scope you discharge the grids of V2 - which have no grid leak resistors. Believe me, they are necessary.

Valves must not be operated with "open" grids. As you have connected them directly to the decoupling capacitors, they build up charge during operation. The data sheets specify maximum values for the resistors (usually 1M for preamp triodes, you will be fine with this. At the moment you are at infinite Ohms).

The gain structure of the whole amp may be something to think about, it might be wiser to drive the active EQ stage with higher level (you lose a lot in the passive tone stack before).
 
Last edited:
I meant "grids of V3" of course... :)

I do not think the cathode follower V4b is necessary. It does not contribute to overall gain and there is no need for buffering.. You could transform it to a standard anode follower stage (probably with unbypassed cathode resistor), this should give you more than enough gain then. But I would possibly change the position of the master volume control to after that stage.
 
Not sure what the purpose of this cathode follower was in the original Hiwatt design. Obviously it what so important to Dave Reeves that he put up with adding a whole twin triode, even leaving one system unused. Mainly, it drives the PI with fixed bias. The Hiwatts of the 1970s even use it without signal going through it, just as a voltage source for the PI.
 
I meant "grids of V3" of course... :)

I do not think the cathode follower V4b is necessary. It does not contribute to overall gain and there is no need for buffering.. You could transform it to a standard anode follower stage (probably with unbypassed cathode resistor), this should give you more than enough gain then. But I would possibly change the position of the master volume control to after that stage.

Now it works properly. I remove the cathode follower, and I changed also the phase splitter for a Fender-like

Thank a lot for your help :)
 
I remove the cathode follower, and I changed also the phase splitter for a Fender-like

Glad to hear that!
Just discovered:
The problem may have been that you used a 100k cathode resistor instead of 220k. As the PI is directly biased by this resistor it probably did not work properly.
Anyway, with a standard long-tailed pair you can't go wrong, although I would consider using a 12AT7 here.

I should do some research on what was the advantage of the Hiwatt design with the voltage source/cathode follower combination. I am curious. Maybe somebody knows?

Cheers,
Georg
 

PRR

Member
Joined 2003
Paid Member
>.....a 100k cathode resistor instead of 220k. As the PI is directly biased by this resistor it probably did not work properly.
> ...what was the advantage of the Hiwatt design with the voltage source/cathode follower combination. ....

Do you mean his R37? 100K here works fine.

I checked it, because we know the Fender 5F6a CF runs on the edge of grid-current and its overload distortion is part of the "sound". But this scheme has the 1.5Meg+1Meg divider in front of the CF. CF grid is down near 100V, and it can easily pull 100K or 220K.

Why so many tubes? For one, I think a lot of guitar design is "try it and see". Some basic plan is adopted from other known-good amplifiers. Then various building-blocks, from other amps or seen in magazines, are tacked-in and tried. For another: sometimes it is about the "looks". The HiWatts are stunningly impressive builds, with their huge chassis and long row of tubes. They may have had a stock of chassis with a lot of tube-holes already punched. Might as well fill up all the holes, give the customer "a lot", even if a more economical design would work the same.

Why did Deusenberg build a V-12 engine? Why did Cadillac trump that with a V-16 (and then another)? Rolls-Royce had shown that a large Six can be smooth enough for any car. (Their later V-8 was more about lower hood than smoothness.) Bugatti Royale made-do with an Eight. But open the hood on a Twin-Six or a Caddy 16, you just say "WOW!" When you are trying to jack-up the selling price to make the most of/off your wealthiest customers, sometimes over-design is good design.
 

Attachments

  • Hi-watt-R37.gif
    Hi-watt-R37.gif
    19 KB · Views: 195
Do you mean his R37? 100K here works fine. I checked it, because we know the Fender 5F6a CF runs on the edge of grid-current and its overload distortion is part of the "sound". But this scheme has the 1.5Meg+1Meg divider in front of the CF. CF grid is down near 100V, and it can easily pull 100K or 220K.

Hi PRR,

you're right of course, it does not make much difference. The determining parts are the voltage dividing resistors at the input of the CF. My bad.

But I think the CF's DC output is way too high. About 100 Volts with 350V supply which should not be too far away from reality! No wonder the PI won't work properly. And not the best idea for the cathode isolation. Change the grid resistor to maybe 220k and things look different.

In the Hiwatt schematic both 12AX7s run on a rather low voltage (with a 100k filter resistor in their HV supply), on the clone probably on much more (especially V4a), with a direct effect on the DC output of the CF. Later designs of the Hiwatt pre dropped the separate filtering for both triodes and lowered the input voltage of the CF instead (which only serves as a voltage source there) by changing the grid resistor to 220k.

But it is certainly ok if it works now.
 
Last edited:

PRR

Member
Joined 2003
Paid Member
....I think the CF's DC output is way too high.....

Ah! You are right. Taking 350V supply, 102V at CF output, 24K in longtail, and 100K for plates, there is only 33V across the one driver, 68V across the other. (They may self-equalize better than that; it is so jammed-up that I don't care to calculate exactly.)

Tail resistor could be 39K or so for better conditions. The OP could actually observe what is going on and ask for data-based comments.
 
It would be interesting to know how much the DC-coupled phase inverter - or the fixed bias circuit in later revisions - contributes to some "Hiwatt sound".
I don't think Dave Reeves did it just for fun, the design can be found in his earlier Sound City amps too. Additional tubes were - and still are - a costly thing, especially in old Hiwatts with their finicky wiring. I don't think musicians could be impressed by an additional tube that adds no real features.
The advantage over a "standard" long tail inverter still is not really clear to me.
Maybe I can find something more on the web.
 
Hi,
I tested the amplifier with more details, and now there is an issue that still pop up. When the amp is switch ON for at least 10min, there is an oscillation PLUK-PLUK-PLUK coming up. I changed the position of every knob to see if there is a correlation between that, and the oscillation. However, there is not.
I also add another filter stage at the power supply, just to check if that could help somehow. No luck so far.
I attached the latest version of the circuit to this post. I appreciate any help that you could get me
Thanks
 

Attachments

  • Circuit.png
    Circuit.png
    180.5 KB · Views: 149
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.