effects loop idea

[daz]

Sounds rather harsh and bright now. The input knob on the DSP has to be almost all the way down. So i think i just need to something along the lines of 10k for the FB resistor on the 1st triode. With a 10k there in series with the mistaken 2.2k it sounded fine and the input knob didn't need to be real low. How do you determine what the DSP requires? The specs just quote the numbers on the switch, +4/-20 dB. If i switch between the mistaken 2.2k thats there and that in series with a 4.7k the difference is a lot. So thats obviously why a 33k is slammin it. I think i may just leave it with the 2.2k since it works fine and sounds good like that and going too high sounds bad. So somehow i guess this DSP requires a much smaller signal to sound optimal.

[Chuck H]

Well I wanted to start with 22k. I only went to 33k because you had low signal to the DSP. Just use 10k and you're good. Don't use 2.2k or you'll have to reduce NFB in the return stage. We want the NFB because it linearizes the stage. So put in the 10k on the send circuit and then compare the treble wiper to the master input.

[daz]

Right now i have about 7k on the send, a 4.7k and a 2.2k in series and the original 220k at the return. I tried adding to the 220k after measuring the pot along with the original 220k but to be honest it didn't sound good. So i sat there with clip leads and bypassed things till i found what sounds best both at low and higher volume. Thats how it is now....7k FBR for the send, (i think i could put 10k there no problem but why?) With a bit over 2vad when i had the 33k there it was slamming the DSP, so even tho it;s only about .8vac as is i think thats what the DSP seems to like. 220k for the other FBR. And i know it's not theoretically correct with a smaller signal to the master then whats coming from the treble pot, but the thing is Chuck, i tried going back and fourth many times with clip leads and it just sounds best were it is now. Cant remember if it was mainly the 220k on the second triode of the one on the send, but increasing one of them just sounded quite bad. It just gets nasty if one of them goes much bigger. Looking at it just from a tonal standpoint I think i should leave it like this because theres plenty of range with the dsp, the amp gets plenty loud, and the volume seems very close to unity if i play the DSP then patch the loop w/o it. I hear no difference in volume to speak of, and the tone is great. So i think i should probably leave it as is even tho it's not theoretically correct. If i ever decide i want to put something else in the loop i can always add more resistance to the 220k FBR. In short, it does everything i wanted it to do as is so i think it's best to leave it as is.

[Enzo]

You know, when you find yourself in situations where a lot of part value changes are happening, why not just mount a little trim pot there. That way you can set up ANY resistance you want in two seconds, and can even swap two values back and forth repeatedly without lifting a soldering iron. You are talking signal path here, so no particular current is involved. And once you settle on a value, well, nothing stops you from just leaving it a trimmer.

[daz]

I do that all the time. But the reason i had resistors in series was this tube addition had no place to easily go ad is squashed in between the preamp tubes and board. And no pace for components on the board. So the entire mess of parts are sticking up from the tube pins and to remove then]m and put a different one in was impossible. I could have used trimmers but there was no point because i was following Chucks instructions on what values to get unity gain and had no idea i;d beA/B'ing values because i thought the correct voltages were all that mattered. Once i got it right it didn't sound good so i started using clip leads to see which added resistors was causing it. I didn't need trimmers because it sounded great and worked great before i added the suggested values. Al i had to do is by pass them to see what was causing the tone to go bad. But i use trimmers and pots all the time like you suggested Enzo. In fact, Chuck had me do that with the return stage to get the return voltage unity with the input. The pot added 86k at unity so i removed it and put a 82k resistor there. But it didn't sound good so i bypassed that and the added resistor to the input side with clip leads sepoerately then at the same time to see what was causing the tone to degrade.

[Chuck H]

I gotta say daz... You do so many things that are either 'something ELSE you did', something you didn't mention or tweak in between checks that it's almost impossible to bead the sight on suggesting anything to you. It doesn't escape me that your SN rhymes with a derogatory moniker

Case in point. What is the feedback resistor in the send stage right now? You didn't say! You added resistance to the return stage and it was ok until you took out the pot and put in a fixed resistor?!? That hasn't happened to me. The wiring catastrophe may be contributing feedback loops (positive and negative) changing every things relativity. That is something that should be eliminated as a factor for the purposes of circuit evaluation. But you continue to attempt circuit evaluation in the face of this circumstance.

I'm not at the end of my rope by a stretch It's always gratifying when clarity is achieved.

I said:

Well I wanted to start with 22k. I only went to 33k because you had low signal to the DSP. Just use 10k and you're good. Don't use 2.2k or you'll have to reduce NFB in the return stage. We want the NFB because it linearizes the stage. So put in the 10k on the send circuit and then compare the treble wiper to the master input.

Since I don't know the current return circuit values we are lost now. Give ME some clarity. I prefer to NOT post arbitrarily.

[daz]

Well, thats just me. I've always been very quick to get things done and once i do i can't just sit and wait for answers if it doesn't work, especially if i'm pretty sure i know what needs to be done. I have to change what i think needs changing. Between that and mistakes i made in the circuit which were mainly due to the tedious nature of trying to put this together in a tiny space, it all leads to all sorts of changes and trying to communicate whats been done via text in forum posts that may be hours apart is extremely confusing to say the least. Not just for you, but for me too. Plus I have a bad memory when it comes to recalling changes and i hesitate at times to even tell you certain things because i'm not sure if i'm recalling them correctly which will lead you to another WTF moment.

But at this point it's exactly as you drew it except for the NFB value at the send because that was one of the mistakes i made initially so i raised it a bit. Tried a larger one as you said but not good, maybe because the DSP prefers about a volt rather than 2.3 or whatever that was? 10k may be ok, but the lower it goes it seems the better it sounds up to a point. What it comes down to is i DID have it *right* theoretically at one point with 15v on both the treble wiper and master input with 2.2v at the send. But it sounded like crap. So at that point it made no sense to be to use theoretically correct values/voltages when it sounded much better with incorrect ones. So i abandoned the "correct" voltages for whatever sounded best and allowed the amp to drive to full volume, or at least as loud as i'd ever need. I don't know if it's the DSP or what, but you gave me the info i needed to make the voltage correct from your vantage point which is one where you can't hear it. And therefore you could only give me the advice i'd need to get the right voltages. So you nailed it. But since you aren;t able to play the amp it was up to me to make the judgment call on the tone, and this is just how it works best.

In any case, you gave me a ton of info that i learned a lot from which i appreciate greatly, Probably more than i've learned in any given 6 month period tweaking amps. And in the end the loop works great. That was the goal and you got me there. Thank you !

Case in point. What is the feedback resistor in the send stage right now? You didn't say! You added resistance to the return stage and it was ok until you took out the pot and put in a fixed resistor?!? That hasn't happened to me. The wiring catastrophe may be contributing feedback loops (positive and negative) changing every things relativity. That is something that should be eliminated as a factor for the purposes of circuit evaluation. But you continue to attempt circuit evaluation in the face of this circumstance.

I tried in post 108 above as quoted below. Maybe i didn't say it in a way that was obvious but if you look at it it tells you what the values are now on the send and return feedback resistors and why. When i removed the pot and installed a series R of the pot value i measured (pot measured 86k, use a 82k) it didn't sound good either with the pot or the resistor. I didn't say it sounded good with the pot and bad with the resistor, i only said the voltages were correct with the pot or the resistor in place to add 80+K. Once it was all done and the voltages were correct i played it and it didn't sound good which led me to use clip leads to bypass the added resistors on both feed back stages.

Post 108:

Right now i have about 7k on the send, a 4.7k and a 2.2k in series and the original 220k at the return. I tried adding to the 220k after measuring the pot along with the original 220k but to be honest it didn't sound good. So i sat there with clip leads and bypassed things till i found what sounds best both at low and higher volume. Thats how it is now....7k FBR for the send, (i think i could put 10k there no problem but why?) With a bit over 2vad when i had the 33k there it was slamming the DSP, so even tho it;s only about .8vac as is i think thats what the DSP seems to like. 220k for the other FBR.

[Chuck H]

Hey daz...

I had to post this because at this stage, this is the "right" way to do it. I know you'll go with your own perceptions, but any lack of correct operation will bite your ass when you need to use the amp in a variable and real world environment. I worked hard to contrive a transparent effects loop that also corrected phase to stuff into an existing amp where NO change was the ideal (if not an ABSOLUTE possibility). I just want the process to be complete in the thread for future readers. Your participation, the quality of your testing and processes and your continued participation isn't within my control. Know this, if this amp were on my bench it would be sounding right and the effects loop would be an invisible addition.

[daz]

I had it like this at one point, except that the send was 2.2v instead of the now recommended 1.5v. I clearly mentioned it a while back. I was getting the same 15v at the input lug of the master and treble wiper. So basically i had it just like you're saying except for the send level. Now the tone sounded harsh. I don't know if it was some mistake i made, maybe something was shorting or who knows what. However, i thought the send was too hot if i recall. Kook back and you may find a post where i said that. Right now i have it similar tho. I abandoned what i last described because i considered what you just said in that last post.....what will happen at a jem or gig. So i just measured it and the send is low as i think i purposely did it thinking that was the reason for the harshness. But the treble wiper and master are close. But i will redo it like this and get them exact or very close and set up the send for 1.5v. I'll post back with results in a while.

[daz]

All i had to do is adjust the send. Worked out to the original 22k. Tried a 10k pot and realized that wasn't enough to get 1.5 at the send which it now has. the master and treble pots are under 1v difference now and i didn't have to do anything there because i already has a second resistor along with the 220k that i determined a while back would give the master matching v in respect to the treble wiper. So in short i just changed the send so it now has 1.5v instead of about 1 v. Things sound the same but it now meets the specs you want so i guess it's done. Thanks.

[Guitarist]

Well, if you had a CF feeding the tone stack and reduced the cathode value say to 22K (but went 3watt resistor) and subbed in a 12DW7 type, and maybe a 100k or 220k pot you'd have a manageable send level. Then you could look at boost options. Maybe a Booste Grande, I hear about. Or you could wire in a tube like a 6u6 (?) that has a pentode + 12au7-type triode. Which will crank your return levels to optimum. A ways back, I referenced a Scott schematic for this.

[pdf64]

Well, if you had a CF feeding the tone stack and reduced the cathode value say to 22K (but went 3watt resistor) and subbed in a 12DW7 type, and maybe a 100k or 220k pot you'd have a manageable send level. Then you could look at boost options. Maybe a Booste Grande, I hear about. Or you could wire in a tube like a 6u6 (?) that has a pentode + 12au7-type triode. Which will crank your return levels to optimum. A ways back, I referenced a Scott schematic for this.

I think that a design objective for an fx loop should be for the send circuit impedance to be reasonably low, in order to avoid the low pass filter (formed in conjunction with the capacitance of the send circuit cabling) reducing system bandwidth noticeably.
I don't think that the above arrangement would readily achieve a low send circuit impedance.
I acknowledge that some other fx loop designs don't achieve it either, including the Dumble http://bmamps.com/Schematics/dumble/loop.pdf (unless the send control is set near either end of its track).
The point being that it's not just about getting the signal level suitable, but also mitigating the potential for signal degradation (eg noise, bandwidth) that the fx loop cabling introduces.

[Tubenit]

There is quite a bit of information on effects loops here: tube effects loop on board

With respect, Tubenit

[pdf64]

From the examples linked in the previous post, I'm baffled trying to work out a technically valid rationale for using a CF buffer followed by a 250k pot for the send control?
The obvious treble loss at mid settings being mitigated by a bright cap; apparently a classic 'there I fixed it!'.
On the face of it, it just seems a unfortunate / uninformed choice that gets unquestioningly copied

[Chuck H]

Indeed! But I see A LOT of DIY circuits like that. Hell, I probably MAKE circuits like that. Well, not like THAT because I happen to know better, but there's still a lot I don't know and I'm sure a technical screening of some of my ideas would reveal it. But yeah, following a cathode follower (for low impedance output) with a 250k pot (back to a high impedance output) is sort of like dieting by splitting your meals into thirds but still eating all three portions

I don't like send/return controls on effects loops anyway. Any decent unit (single effect or "processor") is either set up for unity gain or has controls for mixing on board. From the mid to late 80's there were a lot of amps that had such adjustments and it confused players then as much as it does now. Causing way more problems than solutions. An effects loop should put out about 1.5V at the send and re-amplify 1.5V at the return. Why 1.5V? Just because it's a number that seems to work with the least problems. Obviously "line level" is under one volt, but it's a lot harder to avoid signal degradation and an decreased noise floor processing a signal that small through a guitar amp without added design considerations. All the effects and processors I've tried will handle the 1.5V just fine. Many modern manufacturers effects loops put out a good bit more, but do have occasional problems with compatibility. So it's 1.5V for me until it stops working.

*That's 1.5V at max send output.