New 183 build

[dorrisant]

The power transformer, power tubes input grid resistors and dropping string resistors.

[jazzbass]

Thanks Dorisant,

The Power transformer obviously suitable for powering the 6V6s, the output transformer suitable for the Raa impedance of the two or four tubes, but for the input grid resistors and dropping string resistors what are the values suitable for these tubes.
In which discussion on TAG can I find a comprehensive answer on the subject? I built a Dumble 2nd generation clone with 4x 6V6 with a VOX AC30 output transformer that sounds really good (thanks to erwin_ve and rootz) but I wish I could say it's not just luck but a good build !

Thanks, Franco

[martin manning]

The Power transformer obviously suitable for powering the 6V6s, the output transformer suitable for the Raa impedance of the two or four tubes, but for the input grid resistors and dropping string resistors what are the values suitable for these tubes.

A 100W PT could obviously handle the current requirement for 4x6V6 (~40W), but the output voltage will be high for 6V6, and higher still due to the reduced current draw wrt 6L6. Heater voltage will be high too, since 6V6 use less than half the heater current of 6L6. You could easily bring the plate voltage of the preamp tubes back in line by increasing the first dropping resistor, but the value will depend on where the plate and screen node voltages land with the reduced load. A power transformer suited to 2x6L6, with ~400V B+ would be a better solution. Grid resistor values are not critical, but they do have an effect on bias excursion and the onset of blocking distortion. 1k5 should be fine.

Best to start a new thread on this topic...

[Stephen1966]

If we look at the resistors coming off the PI trimmer in the 183 we can see he has 120k and 100k.

183_side - PI resistors.jpg

I thought I would give that a try. Currently with the 110k/100k we see in the schematics and in Tony's layout, I have the PI trimmer set at at about 10:30 and around a 9V difference between the plates. This gives a balanced signal measured at the power tube cathodes.

With the 120k resistor, in place of the 110k to get the same output the trimmer has to be around 9 o'clock which leaves even less room if I want to reduce the voltage difference between plates to 6.5V/7V.

Playing around with an Amaj. scale I get a nice, sustaining and balanced sound across three octaves when the trimmer is set to provide around 9V difference, but the sustain is lacking when I dial the trimmer back to 7V. The 7V difference, commonly used elsewhere has a marked imbalance at my power tubes' cathodes but it may well work with other tubes. As I remarked myself elsewhere, I've had PI tubes that balance around 9V and others that balance around 2V, so depending on the tube, the voltage difference isn't an especially good standard to set the PI trim by.

The 120k/100k combination then, suggests MrD chose these values according to the tube he was using... I'm assuming the input side was operating at 307V and the follower side at 302V (a 5V difference) as per the layout but if were to take it there, the imbalance would be extreme. TAG (the guy, not the forum) has said the voltages are not correct in the amp garage files but they are probably somewhere close. I'm getting around 300V on the input side and 291V on the follower side. Close enough for Jazz

[bepone]

this PI setting is working together with output tubes, gm, transconductance, so we dont need to measure voltage on anodes of PI, better to scope output currents via out.tube cathodes (1 ohm there or so) instead

[Stephen1966]

this PI setting is working together with output tubes, gm, transconductance, so we dont need to measure voltage on anodes of PI, better to scope output currents via out.tube cathodes (1 ohm there or so) instead

Thanks, yeah! I agree. This is what I kept going back to, the experiment was enough to satisfy my curiosity about the PI plate voltages though. I think I have a good set up now so I've brought it upstairs to just have a play with it. Away from the bench I will be able to run through various levels and settings and let my ears do the work.

[sergio]

Hi , nice build! congrats!

About the 50/100w switch , work fine?

[jazzbass]

Hi , nice build! congrats!

About the 50/100w switch , work fine?

ciao Sergio,

I made a 100/50 switch in a Dumble/Two Rock clone amp, I didn't like the result but by now I had implemented it and it remained. I only use the 100W position.

A hug

Franco

[Stephen1966]

Hi , nice build! congrats!

About the 50/100w switch , work fine?

Hi Sergio. Thank you. The chassis is a bit of a salad strainer with all my experimentation, but I think it came together quite nicely. The 50/100W switch works just fine. Of course it's not half volume, but if you're just practicing away in your bedroom, it's a way to moderate your reputation with the neighbours

Personally, it's not an issue for me, I keep it on 100W. And very nice it is too. Still some twiddling to do but I'm starting to enjoy it.

Tonight, I ran my new pedal in the FX loop. A Hall of Fame 2 reverb. Just great. I built my Skyliner with a 3 spring reverb and that is very nice but the circuit, the reverb pan, and all the rest demands a much bigger chassis. That amp has the FX loop as well and I used it as the base design for this one. I'm very happy with both.

The FX loop is not totally transparent, the cap and pot mods I incorporated after Tony's experimentation and findings were incorporated in this design as well. And (it's hard to describe - but check my previous posts) I think it enriches the tone, in a very positive way. You can - as I did - stay with the stock design as well, and that is even more transparent. There's a lot of latitude in the send and gain pots of the FX loop, and quite easy to find a comfortable medium. If you use time based effects in the slightest, it's a no-brainer.

(Where are you Tony? We miss you.)

The FX loop is a great addition though, and a way to bring this design into the modern era. The modern FX pedals both equal and often exceed the capabilities of more analogue designs, the loop circuit requires some design choices from the ground up and I think in the past, people ran into problems because it was being transplanted on top of the existing designs, but the careful consideration of signal paths and gnds seems to solve those problems. I think it's probably just harder to execute well if you try to add it to an existing circuit/layout.

A dead quiet noise floor. And I really just hammered the reverb pedal tonight, no thought to series or parallel side chaining or optimising the send and gain. I'm just hammering everything, over time, I will dial it in at a more granular level and start to explore the sonic possibilities in (a lot) more detail. Great first steps though.

I think it's almost time to release the schematic and layout. I've held back while I was still figuring stuff out but I think it's all there now. Let me know if you are interested in the build.

[ijedouglas]

The 120k/100k combination then, suggests MrD chose these values according to the tube he was using... I'm assuming the input side was operating at 307V and the follower side at 302V (a 5V difference) as per the layout but if were to take it there, the imbalance would be extreme. TAG (the guy, not the forum) has said the voltages are not correct in the amp garage files but they are probably somewhere close. I'm getting around 300V on the input side and 291V on the follower side. Close enough for Jazz

I think he used a 120K/100K for 2 potential reasons. A) He liked the imbalance it caused B) the output transformer was mismatched and he was using the PI to correct this. I don't believe he did this for a specific tube as this creates a nightmare when replacing this tube. BTW I have seen the exact same 120K/100K in another Dumble. This particular amp is a 50W EL-34 and has a different output transformer (which would make me lean towards reason A unless this transformer was also mismatched)

I have the 120K/100K combination in my 183 clone and really like it.

[pottedplant]

Hey Ian hope you are well, the OT's are physically wound in such a way that stacks one winding on top of the other making the resistances vary for each tube pair (or singular tube in 2 tube amp). From what I gather the amount of turns on each side of the tap must be the same (this is balance?), the resistance isn't as big of a deal and is mostly tolerated or accepted by builders (perhaps Dumble wanted to take it further?). OFC I am no expert, only sharing what I've come across in my journey.

[Stephen1966]

Just an update... Completed schematic and build photos posted in the first post. All done

[Stephen1966]

I've just had a very interesting set up session with the amp and it's sounding glorious. The overdrive is Marshallesque, grainy and wild - how any pedal could hope to achieve these sonic heights, I have no idea! No need for a pedal on 'Satisfaction'. The sound, isn't compressed in the slightest - great articulation/touch sensitivity - but with the bright switch (front plate) in the highest treble setting (set down), the treble turned up to around 1-2 o'clock and both volume and master around midnight, the amp starts with the feedback very nicely. The hot pickup on my bridge helps but it will produce feedback on the neck and mid positions (bridge and neck) as well. The onset of feedback is contained and slow, it can be several seconds between the note fundamental starting to decay and the feedback starting, so it's controllable and the position of the treble allows a certain degree of duration (before onset) control. But put a compressor in front and there would be a ton of sustain. A hypothetical, but likely.

I had one last dive into the bias settings and at 62% it is really juicy. Lower settings, also sound nice, I've had it down to 55% and that sounded good, great even, not at all anaemic but the notes, I get the impression, decayed a mite faster. I also tried it hotter, around 66% but there, the notes started to sound a little sterile so I dialled it back. 62%, comes in around 33mA on the cathode and is on the conservative side and so should prolong the life of the tubes. They're a little hotter than in the original which was biased to around 56.5% (if we use the figures on Tony's layout) but these are JJ EL34s and I don't know what the original used. I understand that JJs respond better when they are a little hotter. Tonally, the hotter bias works best for me though. A nice balance between the harmonic colour of the notes and the sustain. I've also just swapped the JJ 12AX7 PI, for a TungSol 12AX7 and so I'm going to have another look at the PI trimmer but it's really close and I might end up leaving it somewhere near the midnight position where it is now. I like the TungSol more - it seems to offer a bit more headroom. Sustain of the fundamental notes is great though - over half-a-minute - and feedback would literally build and go on forever if I didn't stop it.

Early tuning of the overdrive was giving me a pretty swirly sounding bass. A bit too mushy, but this was with the PI trimmer set to around 30k to ground. That was with the amp cold and not under power. I'm a bit reluctant to test the resistance to ground of the OD trigger with the amp under power but after I adjusted it to 22k ground the notes started to become clearer, more defined and a lot easier to delineate in a chord. It might go a tad higher but this sounds great now so I'm leaving it where it is.

Tuning up in the beginning was really just about getting it in a place where sonically at least, it was more or less okay, but as I get into the final stage now, it's much more about the fine incremental details. I started by saying that the amp is more than a collection of parts, I would go further now and say that it almost doesn't matter what parts are used as long as they are of good quality, it's the build philosophy and the critical ears that are the essential elements of a successful build. There is almost nothing in my amp that was used in MrD's 183, transformers are modern stock, the precision power section uses modern parts, there's a buffered loop in there as well, but having listened to all the 183 clips I could lay my hands on, I am convinced I am pretty close now. I worked on a Marshall DSL 401 recently and the overdrive was really underwhelming compared to this beast. Maybe against a proper high gain Marshall the OD wouldn't have the bite that those Marshalls have. But it's definitely a rocker and yet switch it over to clean settings and it will do BB King. I'm really liking the Sprague Atoms in the preamp's cathodes - the highs are rich and toneful and have just the right amount of presence. The presence pot has barely moved away from the midnight position but when the amp gets in front a band, that will come into its own.

There comes a time when we have to tell ourselves, "it is finished," that's the final challenge. Is anyone still wondering why it sometimes took years for MrD to finish an amp?

[Stephen1966]

The never ending story...

Took the amp to a gig last Friday and it sounded fab. The guy whose Marshall I'd repaired was running his Marshall/Carlsbro stereo stack earlier in the evening and by 1am I was ready to split but I persuaded him to give it a go so we hooked it up to a 15" closed back cab and something amazing happened. First the band came back, then the audience came back, and then the police We had to moderate the levels from then on but it lasted until 6 in the morning when we were all played out. I really expected a 15" closed back cab to sound flabby, but not at all and though the presence did get pushed a little to compensate, what blew my mind was how focused the guitar now was (compared to the Marshall stack), front and centre of the band's sound - if you don't count the singer! Makes a real difference when you can just sit back and listen. The guitar guy was running a stack of pedals into the front of the amp as well and there were some great moments when the sound opened up into 3D.

This came after I had swapped the choke's wires - an experiment I ran, connected with Ian Douglas's great work on his amp: https://ampgarage.com/forum/viewtopic.p ... 81#p451481

I was just chilling out with the amp over the weekend and running off the wall voltages (not the customary variac) as well as using a much shorter speaker lead and for the first time, I noticed a loud switch pop as I put the amp in standby to turn it off. The solution was to put an RC in parallel with the standby switch.

1683035872779.jpg

The B+ is around 460V there so I used R = (V^2)/P to find the value of the resistor. Plugging in the values for a 2W resistor, comes out around 105k so I used a 100k. The cap is a .01uF/2kV ceramic I had left over from the rectifier I didn't build. Does the job perfectly and it shuts off silently now.

Back into playing, I then noticed blocking distortion on the low E (tuned to Eb) notes. This is on the clean channel and the likely culprit is the first stage of V1 which has a grid stopper of 22k. The second stage of V1 has 220k Rs with a 500p in parallel so I'm thinking this is coming from the first stage of the tube. I'm going to think about it some more and welcome suggestions. One solution is possibly to increase the value of the 22k to something like 33k and put a small cap in parallel to maintain the highs. Another, to reduce the value of the coupling cap but I would first have to check if occurs with the R/J switch in both positions to see if that makes a difference. On the other hand, this could be because of the swap I made to the choke leads. I like the kind of blanket tone I got there but that may be the root of the problem and the cleaner, crisper articulation and note delineation Ian switched over to may be for the best. At least I have some ideas of where to look so it shouldn't take long to figure this out.

[bepone]

V1 is not likely to be a problem for "distortion" if is middle biased (see the voltages in cathodes for V1a-V1b),

Probably you are refferencing to too much bass in the clean channel or too much bass going to the PI, where values of some caps need to be reduced.

To attack this, you can lower the value of the cap in the PI (positioned on second grid to gnd) from 0.1uF to 22nF and then check again, and lower values to input to the PI 22nF to 1nF (for checks only)