New 183 build

[Stephen1966]

FET's up and running The bias was way off on B+ after dialling it in using the battery in series. This could be something to do with the addition of the relay and the level pot. Anyway, a word to the wise... if you use relay switching and a level pot on the front panel expect to adjust your bias accordingly. The board with precision trimpots is working just fine.

Also, set the bias to around 30mA which takes the B+ down to about 460V and about 56% plate dissipation. It's pretty cool, in every sense of the word, but this weekend I'll be hooking it up to the scope and tinkering with it a bit more. A great amp, and quiet as a church mouse.

[martin manning]

FET's up and running :D The bias was way off on B+ after dialling it in using the battery in series. This could be something to do with the addition of the relay and the level pot. Anyway, a word to the wise... if you use relay switching and a level pot on the front panel expect to adjust your bias accordingly.

Probably has more to do with the Vdd you have now vs. the 18V battery.

[Stephen1966]

FET's up and running The bias was way off on B+ after dialling it in using the battery in series. This could be something to do with the addition of the relay and the level pot. Anyway, a word to the wise... if you use relay switching and a level pot on the front panel expect to adjust your bias accordingly.

Probably has more to do with the Vdd you have now vs. the 18V battery.

Really? How so? The only thing that makes sense to me is that in the battery biasing method, the earth is tied to the input, creating a virtual earth (please correct me if I am wrong about that). But in the chassis, the earth is bonded to GND 1 (near the input jack) and it's almost certainly a different ground potential. In my battery biased prep (before I adjusted the bias trimmers operating with B+ ~300V) I was getting continuity to ground across all the relay pin outs when on paper it should have been just com1 and nc1 to GND.

IMG058.jpg

This difference in biasing between the battery and B+ isn't a problem I have heard other people, building the more traditional circuit, experiencing. This is why I put it down to the unconventional addition of the relay switching and the external level pot. This is where the trim pots are useful though. Fixed resistors would have entailed removing the board and using trimpots or a decade box to find useful values. I mean, fixed resistors might work fine in the normal set up but if you go the route I went, the trimmers are a godsend. Five minutes, to sort it all out, is all it took.

[martin manning]

Is your FET relay connected as shown here https://ampgarage.com/forum/viewtopic.p ... 08#p365408 in the .pdf titled FET Boost Relay?
What is the voltage at the top of the drain resistor now?

[Stephen1966]

Is your FET relay connected as shown here https://ampgarage.com/forum/viewtopic.p ... 08#p365408 in the .pdf titled FET Boost Relay?
What is the voltage at the top of the drain resistor now?

No, not exactly.

This is what I have...

DELTA 183 - 1.7 - FET.jpg

Voltages are as indicated after adjusting the trimmers (P2 and P3).

DELTA 183 - layout 01 - FET.jpg

[martin manning]

No guarantee that your preamp rail voltage and divider will result in 18V Vdd. If you were able to dial the Vdd back to 18V with the voltage divider trimmer, I would expect the bias to come back as you had it with the battery, without needing to move the source trimmer, since the battery method applies 18V at the top of the drain resistor.

I notice that when you switch your FET boost in you are paralleling 3M3 and 1M resistors, which reduces the input impedance to ~0.7M. In the schematic I linked the FET stage input impedance remains at 3M3.

[Stephen1966]

No guarantee that your preamp rail voltage and divider will result in 18V Vdd. If you were able to dial the Vdd back to 18V with the voltage divider trimmer, I would expect the bias to come back as you had it with the battery, without needing to move the source trimmer, since the battery method applies 18V at the top of the drain resistor.

I notice that when you switch your FET boost in you are paralleling 3M3 and 1M resistors, which reduces the input impedance to ~0.7M. In the schematic I linked the FET stage input impedance remains at 3M3.

One coffee wasn't enough to get my head around this post this morning but a gallon of coffee later, I think I understand the problem, where it originated, and how to fix it.

First, thank you Martin.

The problem started when I moved away from separate jacks for FET and NOR to a single jack input. The original two jack arrangement effectively places the 1Meg grid leak resistor after the FET circuit. By leaving the 1M in place on the jack, when converting to a single jack it sits in front of the FET circuit and yes, in parallel with the 3M3 resistor. So what you did, is to relocate the 1M after the FET, just before the grid stopper of V1's grid. Clever stuff! Having seen the problem, the next problem was how to fix it using the existing relay board topology.

This is a before and after...

IMG059.jpg

I think this will solve the problem. It might even serve to clean up the sound of the FET because at the moment there is an element of distortion and it figures the circuit is having to work harder to process a weaker signal, and that also introduces noise. It's easy to forget that the grid leak resistor plays an important role in biasing the signal between grid and ground and so when I make this correction I wouldn't be surprised to find it's necessary to rebias the circuit again. The Vdd, the Vs (or both) may shift with the new arrangement. Unfortunately, I've lost my original bias settings from the batteries in series set up so I won't be able to say for sure if this accounts for the initial (significant) difference between battery and B+ bias points, but it might be possible to get an idea of that. I'll make the mod today and report back.

Edit: actually... thinking about this... If I return my Vdd to ~6V and Vs to ~1V that will get my bias settings pretty close to the state they were in after the battery biasing. Probably safer to do that first before re-configuring the circuit.

[martin manning]

It's easy to forget that the grid leak resistor plays an important role in biasing the signal between grid and ground and so when I make this correction I wouldn't be surprised to find it's necessary to rebias the circuit again.

It's ground referenced either way, so the bias point should remain the same.

[Stephen1966]

It's easy to forget that the grid leak resistor plays an important role in biasing the signal between grid and ground and so when I make this correction I wouldn't be surprised to find it's necessary to rebias the circuit again.

It's ground referenced either way, so the bias point should remain the same.

You were right, naturally. In the end, I kept the connections as in the first diagram immediately above, just relocated the 1M resistor to the GND and V1 (grid) lugs. I didn't want to set anything on fire so before I made the mod, turned down the Vdd trimmer to ~6V. Afterwards, when I biased it back up to 18V the Vs trimmer was pretty much on the money. This is a distinctly cleaner sound now, adding a pleasant fullness to the note. Later on, I will try some of the different bias points as you did. I'm looking for a clean boost, and so this is probably pretty close.

[Stephen1966]

Wow! This is starting to sound great.

I reconfirmed the bias today and it sits around 55% with 30mA on the cathodes, calculated by the transformer resistance method (voltage drop between CT and plates).

Next, I started on the PI trimmer

A couple of shots here...

SDS00001.jpg

SDS00002.jpg

This was confirmed later by a listening test, moved the trimmer back a little and the notes started to bloom very nicely. I would spend more time with it though, this is very close though.

Then, I thought I would have a look at the output. These are both at the onset of clipping and it takes quite a bit more before the x-over notch appears but by that stage my load was starting to heat up quite significantly and it's a 100W load.

SDS00004.jpg

SDS00005.jpg

To the listening test after dialing in the PI. The cleans are rich and full even with the mid and rock switches off and zero treble boost (the high switch is a three position with the center cancelling the high frequencies). Just as clean as it could be. A hint of break-up when you dig in. Then, with PAB the harmonic depth is even more so. Then adding the mid and rock switches, more treble and finally OD. The note sustain is especially pronounced here and there is a gritty bite which with a hot bridge is edging towards a scream without the brittle, dry ice-pick sound I get with my solid state amp. Next thing on my list though is to have a play with the OD trimmer because it seems to be a bit thin with the low end compared alongside the clean settings at the same volume. At the moment, it's not been touched and is still in the 12 o'clock position. The sustain is great and the note goes through several harmonic profiles before it finally starts to feedback. The feedback is quite late in developing, which is okay. One thing at a time though

I mean to get it out of the workshop now to put it in a listening environment and try it with a few pedals, try out a few licks and dial it in a bit better than is possible in the confines of the workshop. With it out of the working space I'll be able to get to work on the pedal and the head cab. That's the plan anyway.

Here's a gut shot as of today...

1676837333009.jpg

[dorrisant]

Very nice looking build! I bet you are a proud papa.

[bepone]

very nice.. also sound description is interesting..

[Stephen1966]

Thanks guys! It's close but not quite there yet.

Here's an interesting post: https://www.thegearpage.net/board/index ... t-29569829

I think the cool bias around 55% Pd is about right. I've been looking at my voltages as well and everything is a bit low. I think I might try replacing that 4k7/5W with the 3k again and see if that brings the voltages back up - it will of course. I'm thinking the preamp and PI need a bit more juice. That should also give me a bit more headroom as well.

Here are the bias and voltage measurements from yesterday:

Bias and voltage measurements 19-02-2023.jpg

[Stephen1966]

Today, I swapped the 4k7/5W resistor back to the 3K. I think this has improved the voltage spread across the preamp. Can't remember why I thought the 4k7 was a better choice but it was probably before I sorted the rest of the dropping string out, and that would throw the voltages off upstream as well. Anyhow, it is fine with the 3k, no need to order any other intermediate values...

I also swapped tubes V5 and V7 but I might swap them back because the measurements after the swap show the first order was a better match.

After rebiasing, the PI needed retrimming as well but it will change again if I swap the tubes back - so being late here as well, I'll look at this again later in the week when the neighbours aren't going to have any excuses for complaining.

Here's today's bias and voltage measurements...

Bias and voltage measurements 20-02-2023.jpg

[jazzbass]

Hi everyone,

Sorry for the trivial question but to use 4 x 6V6 as power tubes instead of 4 x 6L6 as well as the different output transformer which other components will I have to change to exploit the tubes in the best way?

A Hug, Franco