Behringer VM1 Vintage Time Machine

A forum devoted to mod, tips and suggestions for upgrading and rehousing your VERY CHEAP commercial stompbox to near boutique excellence.
User avatar
j_flanders
Information
Posts: 16
Joined: 14 Apr 2014, 14:10
Has thanked: 2 times
Been thanked: 7 times

Post by j_flanders »

Some more frequency responses.

The dry signal as it arrives after the compressor, just before the anti alias filter and the BBD section for the Deluxe memory man:
(Level and mix is set for unity gain, feedback is at 50%)
DMM_INPUT_TO_AAFILTER.jpg
Same thing but for the VM1:
The VM1 delivers a much more scooped dry signal to the bbd section. The main 'culprit' is that small 22k bottom resistor (compared to 100k in the DMM) in the voltage divider after the first opamp.
VM1_INPUT_TO_AAFILTER_FB50.jpg
Another thing to note is that, for the feedback pot, because of its large size (50k vs 10k in the DMM), its position affects the frequency response quite a bit.
In the DMM the feedback pot is 10k, which is 'insignificant' compared to the 100k voltage divider resistors, so the postion of the feedback pot has little to no effect on the frequency response.
Here's the frequency response for the VM1 with the feedback pot at 0% or 100%¨, giving an even more scooped dry signal:
VM1_INPUT_TOAAFILTER_FB0.jpg
In the Boss DM2 or 3, from which the feedback path was copied, the small 20k resistor made sense because they use a different way of summing dry and delay signal for the feedback.
The DMM uses the same passive mixer for the feedback as it does for the output mix, but with two 100k resistors instead of the 10k pot at the output.
The DM2/3 uses an active inverting, summing opamp mixer and it actually uses the internal opamp of the compressor for this. It has a 20k input resistor (R3 at pin 6) and by connecting the feedback signal through a 20k resistor to pin 6 you get a 50/50 suming.
But combining the 100k resistor of the DMM with the 20k resistor of the DM2/3 makes little sense.
Compressor internals (pin 3 is 20k, Boss connects feedback signal through 20k to pin 5):
COMPRESSOR_VM1.png
Because the 3205 bbd chips have less headroom and worse frequency response for higher frequencies than the 3005, this 'mistake' kinda works out in the end. The VM1, because of the scoop, sends less signal to the BDD and has a pretty big pre-emphasis on the highs.

But for now we've only looked at the 'mistake' from the dry signal perspective. Next post is about how it affects the delay signal.

User avatar
j_flanders
Information
Posts: 16
Joined: 14 Apr 2014, 14:10
Has thanked: 2 times
Been thanked: 7 times

Post by j_flanders »

Frequency reponse of the delay / feedback signal. Level and mix set for unity gain, feedback at 50%.
In real life the signal coming out of the bbd section has been filtered by the anti-alias filter and the reconstruction filter, but to see the effect of the differences between the DMM and VM1 feedback loop I'm using a FRFR input signal.

DMM: ( - 3dB at 1kHz)
DMM_FEEDBACK_TO_AAFILTER.jpg
VM1: ( + 3db at 1kHz)
VM1_FEEDBACK_TO_AAFILTER_FB50.jpg
The DMM has an extra LP filter, right before the feedback pot.
The VM1 is flatter and has stronger repeats in the feedback. It should oscillate earlier on the dial than the DMM.
The VM1 has a softer first repeat but relatively stronger later repeats.
From the plot you'd expect much brighter repeats on the VM1, but in a later post you'll see that in the VM1 the signal looses more highs along the way(in the filters, bbd's, make up gain etc) than the DMM and will eventually have less highs than the DMM.

Once again, you could say that the 'incorrect' 22k and 100n in the feedback loop kinda works out in the end.

Replacing the 22k with a 100k pot and the 100n with a switch to various caps, really lets you shape the repeats both decay-wise as frequency wise.
22k and 100n is better for faux reverb
100k and 22n is better for distinct repeats.

User avatar
JiM
Diode Debunker
Information
Posts: 928
Joined: 11 Mar 2008, 23:56
Completed builds: Completed builds :
Proco Rat
MXR MicroAmp in a volume pedal
TubeDriver (w/ NoS russian tube and big muff tone contol) + Phase 45 (w/ univibe cap ratio)
Dallas Rangemaster (w/ noisy OC75, negative ground)
SubCaster tube booster (w/ NoS russian tube, PtP)
Hot Harmonics
Music From Outer Space SubCommander in progress
Crackle Not OK
Simple bass blender in a 1590A
Bazz Fuss with a photo-darlington
Location: France
Has thanked: 30 times
Been thanked: 31 times

Post by JiM »

Thanks a lot, j_flanders !
That's amazing work in analysis, error-finding and circuit improvement. :applause:

I'm proud that 10 years later, this thread still exists and finds some use.
j_flanders wrote:Replacing the 22k with a 100k pot and the 100n with a switch to various caps, really lets you shape the repeats both decay-wise as frequency wise.
22k and 100n is better for faux reverb
100k and 22n is better for distinct repeats.
Hmm, that would be a very useful and simple mod. A SPDT switch selecting "bright" or "dark" repeats.

Here is a new version of the schematic, including fixes for the errors you've found. Let me know if there is still something wrong.
Attachments
Fixed thanks to j_flanders
Fixed thanks to j_flanders
I only give negative feedback.

Mr. Shred
Information
Posts: 5
Joined: 18 Aug 2013, 20:57

Post by Mr. Shred »

Hi guys!

I want to mod the Vintage Time Machine to true bypass but can’t find the instructions anywhere. Would anybody happen to have the instructions or a schematic on how to do it?

User avatar
JiM
Diode Debunker
Information
Posts: 928
Joined: 11 Mar 2008, 23:56
Completed builds: Completed builds :
Proco Rat
MXR MicroAmp in a volume pedal
TubeDriver (w/ NoS russian tube and big muff tone contol) + Phase 45 (w/ univibe cap ratio)
Dallas Rangemaster (w/ noisy OC75, negative ground)
SubCaster tube booster (w/ NoS russian tube, PtP)
Hot Harmonics
Music From Outer Space SubCommander in progress
Crackle Not OK
Simple bass blender in a 1590A
Bazz Fuss with a photo-darlington
Location: France
Has thanked: 30 times
Been thanked: 31 times

Post by JiM »

It's just a matter of re-wiring the connectors between the jacks PCB and the footswitch PCB. Referring to the schematic above :
- disconnect X6 pin 1 and X7 pins 3, 4 and 5 : this disconnects the tone-sucking input and the LED control.
- connect X8 pin 1 to X1 pins 3 and 4 : this connects the input to the switch formerly used for the LED, and to the direct bypass path.
- connect X1 pin 5 to X7 pin 3 : this connects back the input to the effect through the "new" switch, therefore only when the effect is engaged. No more tone sucking.

If you are careful, you can remove the crimped terminals from the female connector housings with an pointy tool. Otherwise just cut the wires, tin, put heatshrink, solder, shrink.
To identify the connector pins, note that pin 1 has a square solder pad.

Or a quick search could have given you the answer that is already available for more than 3 years : viewtopic.php?f=11&t=28117
I only give negative feedback.

Mr. Shred
Information
Posts: 5
Joined: 18 Aug 2013, 20:57

Post by Mr. Shred »

Hi JiM,

Thank you very much for your help, I think I'm starting to get my head around it now. Do I lose the led indication in the process, as was mentioned in the post you linked?

I'm thinking about removing the original footswitch and putting in a replacement 3PDT footswitch wired for true bypass like this:

Image

So I'm guessing I should maybe wire X7 pin 1 to effect send, X7 pin 2 to output, X7 pin 3 to input, X7 pins 4 & 5 to the resistor & led lug, and add a new wire for the ground lug. How about the effect return then, maybe something to do with X6 pins 5 & 6? Or would this be anywhere near correct at all?

User avatar
JiM
Diode Debunker
Information
Posts: 928
Joined: 11 Mar 2008, 23:56
Completed builds: Completed builds :
Proco Rat
MXR MicroAmp in a volume pedal
TubeDriver (w/ NoS russian tube and big muff tone contol) + Phase 45 (w/ univibe cap ratio)
Dallas Rangemaster (w/ noisy OC75, negative ground)
SubCaster tube booster (w/ NoS russian tube, PtP)
Hot Harmonics
Music From Outer Space SubCommander in progress
Crackle Not OK
Simple bass blender in a 1590A
Bazz Fuss with a photo-darlington
Location: France
Has thanked: 30 times
Been thanked: 31 times

Post by JiM »

Mr. Shred wrote:
28 Jul 2020, 11:36
Do I lose the led indication in the process, as was mentioned in the post you linked?
Yes, the wiring i've devised yesterday is indeed identical to the previous post. It's only logical.
Mr. Shred wrote:
28 Jul 2020, 11:36
X7 pin 1 to effect send,
Nope, this is the output of the effect.
Mr. Shred wrote:
28 Jul 2020, 11:36
X7 pin 2 to output,
That would work.
Mr. Shred wrote:
28 Jul 2020, 11:36
X7 pin 3 to input,
The input jack is still connected internally to the input of the effect, resulting in tone-sucking in bypass. You need to break that connexion to achieve true bypass.
Mr. Shred wrote:
28 Jul 2020, 11:36
X7 pins 4 & 5 to the resistor & led lug,
The resistors and LED are already on the board, you just need to short those two pins.
Mr. Shred wrote:
28 Jul 2020, 11:36
add a new wire for the ground lug.
This one is intended to complete the LED circuit when on, and to ground the input of the effect when off (as it's disconnected from the input jack).
Mr. Shred wrote:
28 Jul 2020, 11:36
something to do with X6 pins 5 & 6?
Those are for the modulation speed (Chorus/Vibrato switch).

Why don't you just draw the true-bypass wiring previously described on two poles of your new switch, and then on the third pole replicate the LED part from SW1A ?
I only give negative feedback.

Mr. Shred
Information
Posts: 5
Joined: 18 Aug 2013, 20:57

Post by Mr. Shred »

Thanks for the feedback.

So let me try this again:
- disconnect X6 pin 1 and X7 pins 4 & 5
- X7 pin 1 to effect return
- X7 pin 2 to output
- X7 pin 3 to effect send
- X8 pin 1 to input

Then install a new led in series with, for example a 1.8k or a 2k resistor and a new ground wire to the 3DPT switch. Would this work?

Post Reply