5-9-10W Practise Amp with LND150 & ECC99

[Bluemeany]

Here's something for anyone who needs a low volume (although not that low - it'll still loud enough to make your ears sting!) practise and recording amp.

It's got an interesting phase splitter design and an output stage that will deliver 5W clean and 9-10W @ 10% THD.

{EDIT} Slight modification to the output stage - a cap was removed that is no longer in the final build!

{EDIT2!} Swapped around BASS, MIDDLE notes and added Q1 grid resistor.

{EDIT3!} Updated power supply drawing to show correct use of bridge rectifier.

[tubenit]

That is a VERY interesting amp!  THANK you for sharing your idea! 

Any chance that you have any soundclips of the amp?  I'd love to hear some.

With respect, Tubenit

[Bluemeany]

Hi Tubenit, thank you for the kind words. I don't currently have sound samples - I will try to get some rudimentary ones  recorded next time I have it set up. In terms of tones available from the amp you're looking at everything from a nice spanky clean right through to a classic rock. You've got the master volume in there to let you bias the clipping to more pre or power amp or just run the thing flat out for fun.

The one thing that I find sets this amp out tonally from other 5W practice rigs it the fact it uses a "proper" output transformer rather than one of those nasty overwound little things which are usually employed with all their shortcomings.

[PRR]

Lovely drawing.

I would expect a "grid" resistor on Q1?

Tonestack appears to be near-standard Fender 3-knob, but Bass and Mid notes are swapped.

Interesting that you take your first stage B+ at a higher level than the middle stages. There may be good engineering here. I had never seen it done.

Eyeballing ECC99 curves, I suspect the un-clipped power output is 4 Watts. Of course for guitar I have no problem saying 5W clean(ish) and double that when blasting. Especially since it leads to a nifty name.

[kagliostro]

Very interesting solutions on this amp 

Time ago there was a search for a mosfet PI on DIYItalia, thanks for sharing the project !

---

Which are the councils about the PT to be used and, can you specify better the way to achieve the transition from 4w to 9W ?

Thanks

Franco

[Bluemeany]

Thank you PRR, I have updated the points you mentioned above; they were, of course, errors!

The reason for the HT supply being arranged as it is, as I'm sure you can tell, is to allow the maximum possible headroom on the first stage. I was going to use a high voltage regulator to clean this arm of the power supply to ensure lowest noise but I settled with the single RC filter after deciding the benefit from the regulator was minimal.

The 5W clean measurement I'm getting is because I'm running slightly over 400V (420!) on the anodes.... The datasheet suggests a 400V max so that's what I've advised others to do here.

Hi Kagliostro - Thank you for you kind comments. Any output transformer which presents a 6k5 anode to anode load with an 8ohm secondary load will do (it must have a sufficient power rating  of course!), you load this secondary with 32 ohms to get the required primary load. These are standard values for 6L6 output transformers on many Fender designs so people may have them lying around. The power output goes from 4-5W when playing clean up to a maximum of 9W when you've got the amp cranked. That's loud when you're using sensitive speakers such as the G12H Anniversary.

Expect some more mosfet phase inverters in the near future. I've got another project on the boil!

I must reference Merlin's excellent book on preamp design here - anyone who's read it will see the influences!

[kagliostro]

Many thanks for the quick reply 

I was asking about the PT not the OT (about which I've read in your previously explanations)

---

So you mean that the output variation on power is due to the input signal, oh, I see, I was thinking you refer to a variation due to different settings on the bias and B+ voltage

Franco

[Bluemeany]

Hello Franco - I must apologise for not reading your question correctly.

Yes, you are correct, the output power available from this power valve setup and configuration is around 5 watts clean (1-3% THD), as you crank the input signal up the power produced will go up to around 9w but the THD will increase upto 10%+. That's a very noticeable level of clipping. The zener diodes placed at various points around the circuit help to limit signal overload conditions in the phase inverter and output stage. By limiting the signal overload to a predefined level you achieve a nice smooth overdrive from the output stage and avoid any blocking distortion or excessive crossover distortion.

When I manage to get some time back at the board I'm going to be looking at a different phase inverter configuration and possibly driving the output stage into a near class AB2 setup. I have no idea how the ECC99 will take this setup though so there could be some trial and error before I manage to arrange something sensible.

[kagliostro]

If I'm not wrong you have a PT with near 280v AC, correct ?

--

When I manage to get some time back at the board I'm going to be looking at a different phase inverter configuration and possibly driving the output stage into a near class AB2 setup.

Interesting, if you do that, please post your comments about

--

The actual board is a PCB or a Turret/Eyelet Board ?

Ciao

Franco

[Bluemeany]

If I'm not wrong you have a PT with near 280v AC, correct ?

--

When I manage to get some time back at the board I'm going to be looking at a different phase inverter configuration and possibly driving the output stage into a near class AB2 setup.

Interesting, if you do that, please post your comments about

--

The actual board is a PCB or a Turret/Eyelet Board ?

Ciao

Franco

The prototype is build on a mixture of turretboard (plank of thin wood & screws!) and veroboard for the solid state stuff.

[printer2]

If I'm not wrong you have a PT with near 280v AC, correct ?

--

When I manage to get some time back at the board I'm going to be looking at a different phase inverter configuration and possibly driving the output stage into a near class AB2 setup.

Interesting, if you do that, please post your comments about

--

The actual board is a PCB or a Turret/Eyelet Board ?

Ciao

Franco

The prototype is build on a mixture of turretboard (plank of thin wood & screws!) and veroboard for the solid state stuff.

We won't judge.

[Bluemeany]

If I'm not wrong you have a PT with near 280v AC, correct ?

--

When I manage to get some time back at the board I'm going to be looking at a different phase inverter configuration and possibly driving the output stage into a near class AB2 setup.

Interesting, if you do that, please post your comments about

--

The actual board is a PCB or a Turret/Eyelet Board ?

Ciao

Franco

The prototype is build on a mixture of turretboard (plank of thin wood & screws!) and veroboard for the solid state stuff.

We won't judge.

It looks a bit like Frankenstein's monster...

[kagliostro]

But it is alive !


Franco

[printer2]

The prototype is build on a mixture of turretboard (plank of thin wood & screws!) and veroboard for the solid state stuff.


My prototype.

[Bluemeany]

Looks great.

The way I like to look at it is if I can get the build stable under these conditions then making a stable amp in chassis with shielding e.t.c should be much easier. Building like this really teaches you layout!

[printer2]

The prototype is build on a mixture of turretboard (plank of thin wood & screws!) and veroboard for the solid state stuff.

My prototype.



Oops, quote in the wrong place.

Just showing the OP that pristine amps are nice to look at but but that is the final incarnation. They can start pretty rough.

[Bluemeany]

I like the way screws and plywood allow you to quickly put together working turretboard layouts without having to get the drill out. You just need a screwdriver and a set of 10mm screws and you're off.

[Paul1453]

I like the way screws and plywood allow you to quickly put together working turretboard layouts without having to get the drill out. You just need a screwdriver and a set of 10mm screws and you're off.

I also like the way you have L brackets for mounting your jacks/pots/tubes.  Good idea!

Now they told me I had to ditch the little proto boards for such HV tube amp circuits.
I see you seem to be making things work using them.  I have been trying to move away from them.
But connecting that number of components without them turns into a big mess quickly.   

[printer2]

I like the way screws and plywood allow you to quickly put together working turretboard layouts without having to get the drill out. You just need a screwdriver and a set of 10mm screws and you're off.

I also like the way you have L brackets for mounting your jacks/pots/tubes.  Good idea!

Now they told me I had to ditch the little proto boards for such HV tube amp circuits.
I see you seem to be making things work using them.  I have been trying to move away from them.
But connecting that number of components without them turns into a big mess quickly.   

Just have to be mindful of the voltages between the traces. Higher voltages require a trace or two between parts. The wires to the tube sockets are the real concern. Could not do it this messy with a higher gain amp. Still can be done but the excess lead length would need to be trimmed off.

[Paul1453]

In the lower right corner, is that a little DC-DC step up transformer?

It doesn't quite look like this one on my Ebay watch list:

http://www.ebay.com/itm/331591952928

I've been thinking about matching one of these with a Dell 19V laptop PS that has a messed up plug end.

I'm thinking it would work for my little 2 tube 250-275VDC 12AX7 (6BQ5/6AQ5) circuits.

That would be lighter and easier to mount in small 10W SS amp housings.

I'd still need to work out how to use that 19V for both the heaters and the HV.

Now I'm wondering if that heater current would end up overtaxing the Dell PS.

No point in using one of these and also needing a 6.3VAC PT too.

[Bluemeany]

In the lower right corner, is that a little DC-DC step up transformer?

It doesn't quite look like this one on my Ebay watch list:

http://www.ebay.com/itm/331591952928

I've been thinking about matching one of these with a Dell 19V laptop PS that has a messed up plug end.

I'm thinking it would work for my little 2 tube 250-275VDC 12AX7 (6BQ5/6AQ5) circuits.

That would be lighter and easier to mount in small 10W SS amp housings.

I'd still need to work out how to use that 19V for both the heaters and the HV.

Now I'm wondering if that heater current would end up overtaxing the Dell PS.

No point in using one of these and also needing a 6.3VAC PT too.

You could certainly use one of those for valve based circuits - you could use a series resistor or a simple regulator (LM317 types e.t.c) to supply the heaters. If you power the heaters at 12.6v you're only going to need 150mA for each valve, most laptop supplies are in the 2-3A range (well, the hulking old laptop I have has a 3A supply!).

[kagliostro]

One alternative will be to use the ECC99 at 12.6v (400mA) and the ECC83 at 6.3v (300mA)

then connecting the tubes in series will give 18.9v (12.6v + 6.3v)


Franco

[sluckey]

If you will run tube filaments in series the tubes must require the same current.

[kagliostro]

Yes, that is right, a resistor must be added to perform the in-series connection, unfortunately I had some indication about on my archive but

I'm not able to find or remember it at the moment

EDIT: Also adding some diodes in series putting it in parallel with the heaters of each tube is a way to use tubes planned for parallel connection in series, but my bain refuses to remember more

Franco

[Paul1453]

I guess I'll just have to buy one and test it out.  For less than $9 including shipping, it costs less than most of the tubes we use.

Most of the crappy little SS amps I've seen already have a 12VDC supply in them.

If with a 12VDC input, that little step up transformer could put out 300VDC @ 100-120mA that would work well.
I might even be able to go up to a 6AQ5 PP output circuit with about that amount of current.

That would also eliminate the need for a laptop power supply unit. 

[kagliostro]

Well, I've find something, but still not all what I need

Two tubes with different consumption heaters can be connected in series using a resistor paralleled to the tube that has the lower consumption

We must consider the consumption of the two tubes and subtract the consumption of the tube that has less consumption from the oter tube

as an example a 6V6 (6.3v - 450mA) and a 12ax7 (connected for 6.3v - 300mA)

so 450mA - 300mA = 150mA, this is the current consumption to be compensated

6.3v / 0.15A = 42R that is the value of the resistor we must use in parallel with the 12ax7 heaters to compensate the different consumption

the value in W of this resistor can be calculated, 0.15A x 42R = 6.3W

so a 42R - 10W can be used to "balance" the consumption

What I've not found is in relation of the fact that as to compensate the high voltage of the supply one tube is used connected for 12.6v and the other connected for 6.3v and this confuses my poor knowledge

Franco

[Paul1453]

If I could just go with the 12VDC supply already in a crappy SS amp.

Wouldn't it be simply enough to power the 12AX7 heater with that, and just add a resistor in series with the 6AQ5 to drop the voltage to about 6V for that tube alone.   

[kagliostro]

If you prefer you can do that, only the drop from 12v to 6.3v is to be considered as dissipation, do your math and establish which solution represent less waste of energy (I don't know)

--

I've find the news I was looking for, see the attached image

(there is a 10R thermistor to "protect" the heaters, the value is arbitrary chosen near the value of the resistance of the filament)

Franco

[Paul1453]

You are right.  The diagram you just posted is very helpful!   

Just a guess, but I get the feeling my series/dissipation method is going to just waste significantly more of the available current and turn it into heat.   

PRR just posted this on another thread.
He has a way of explaining things that a simple Army trained repair tech like me can easily understand.

> Will this PT's current   also be affected by the resistor that I used?

No. Current is in series. You are just losing voltage to your amplifier.

My house has 125V power at the street. Good for 100A. Glenn put the house way back in the woods, and didn't spend many-thousands for the power wire. I can still pull 100A, if I wish. Voltage at my house will be under 90V, lamps would be dim, but neither the transformer nor my wire would mind. <

Now I'm thinking my dissipation method won't use any more of the available current.   

[printer2]

The 12V dc units usually put out that, 12V or slightly less. Still within 5% of what the tube wants. The cheap ones from China (where else?) can be short on current running on 120V. They put out their rated current at higher voltages (220V-240). The ones with a name brand on them are designed better and get their rated current. I make sure they are rated for more than I want, derating them by 50%.

They make more sense when you are feeding 12V tubes like the 12AQ5's, others I have that are candidates for these little supplies, 12AB5, 12V6, 15CW5. Many other tubes that don't have 6.3V heaters that go cheaper than their 6.3V cousins. I got a lifetime supply of 12V6's for about $5 each. Need to build an amp with them soon, maybe the next one.

[Paul1453]

Good points to consider, printer.

I think I have a couple of the 12AQ5s, not nearly as many as my 6AQ5s.
I think I have a couple octal 12L6s too.

I had not even pulled those out and put them with my more actively used tubes.
I might have put them with the 50 and 35 V audio tubes, that I was not likely to use.

[jazbo8]

an output stage that will deliver 5W clean and 9-10W @ 10% THD.

Nice drawings! I have a question on the 9-10W output, is that measured or calculated?

[Bluemeany]

an output stage that will deliver 5W clean and 9-10W @ 10% THD.

Nice drawings! I have a question on the 9-10W output, is that measured or calculated?

Thanks Jazbo. That figure is measured. I still have a few questions inside my head regarding endurance testing and this setup - i.e there's a few things I could add/do to bulletproof this thing; zener diodes on the source followers to protect the gates and subjecting the whole amp to a 24 hour guitar loop soak test just to see if any problems rear their heads. Anyhow, that said - the amp's still alive and well so....

I'm making progress on the AB2 (ish) version. The phase splitter and mosfet drivers are done as is the power supply. I've got it running now it's just a case of reigning in the figures until I get something usable and reliable... I'm going to be honest - I have zero class AB2 design experience, I've got a general idea of what I'm looking for and I've read the literature that I could find but if anyone here has a design (and measurement) process drawn out and tested I'd love to hear it.

[printer2]

Good points to consider, printer.

I think I have a couple of the 12AQ5s, not nearly as many as my 6AQ5s.
I think I have a couple octal 12L6s too.

I had not even pulled those out and put them with my more actively used tubes.
I might have put them with the 50 and 35 V audio tubes, that I was not likely to use.

I figure when I run out of 12AQ5's (doubt it) I can always wire two 6AQ5's in series. With dc no noise penalty.

[jazbo8]

That figure is measured.

Great, that's quite good from these little buggers! Also, you are not going to get very far with class AB2 with the existing driver stage, for a quick tutorial on how to do it, you can search for Tubelab's PowerDrive circuit.

[DummyLoad]

IMO, a 2A SB fuse is a lot of fuse for a device that requires less than 50VA at full power. 1 amp would be more than sufficient.

--pete

[Bluemeany]

That figure is measured.

Great, that's quite good from these little buggers! Also, you are not going to get very far with class AB2 with the existing driver stage, for a quick tutorial on how to do it, you can search for Tubelab's PowerDrive circuit.

Thanks for the heads-up Jazbo, I'll check that article out. However I've already completely redesigned the driver stage. I'll draw up a schematic later but it's a cathodyne setup using the LND150 again driving two IRF830 source followers taking their power from a bipolar supply. The key to getting this running nicely (by nicely I mean not setting on fire) is going to be selecting a bias point and grid stopper values for the power valves. I had it set up with a differential probe looking at the grid of one power valve and how much current it was pulling - it's surprising how much they want!

I decided to "experiment" and stuck in some 470 ohm grid stoppers. With a HT supply of 400V and a bias of -19v the power amp was developing 10-11W of power before hard clipping sets in. I daren't run this set up for long - constant sine wave testing of AB2 isn't nice for the power valves.

For now I settled on 1k grid stoppers, as I mentioned above this will probably change with further testing. With the same 400v HT and -19v bias I was getting around 7W "clean", this will be more than enough - the main things I'm keen on harnessing are the soft clipping and anti-blocking features the mosfet drive allows me.

The use of a cathodyne setup in the phase inverter has also forced me to redesign the preamp section - I'm going to be looking to achieve ~30v peaks into the cathodyne - that'll give me a nice degree of power amp clipping without driving things to hard. The beauty of the phase inverter design is that it can produce these high voltage swings without adding any of it's own clipping - it'll still be a full valve clipping amp....

So things thus far are looking promising - it works and hasn't exploded yet! Another few days next week fine tuning the power amp and I'll be onto the redesign of the preamp stage. I'm thinking I may allow myself another preamp valve - with 4 individual gain stages I'll have ample tools to shape the gain and frequency response to my requirements.

[Bluemeany]

IMO, a 2A SB fuse is a lot of fuse for a device that requires less than 50VA at full power. 1 amp would be more than sufficient.

--pete

The fusing in the schematic is just what's in my test rig - I've not done full current draw readings on the amp yet. As you suggest the lowest possible usable value is what you should select..

[jazbo8]

For now I settled on 1k grid stoppers, as I mentioned above this will probably change with further testing. <snip>

For class AB2 operation, you usually want to direct coupled the driver (the cathodyne in this case) to the output tubes, so the grid stoppers are just going to get in the way. Also take a look at the Dumble SSS, and the Ampeg SVT for some guidance on how they drive the output tubes, but be sure not to exceed the maximum heater-to-cathode voltage.