Rockman X100 - Make it a DIY Pedal

All about modern commercial stompbox circuits from Electro Harmonix over MXR, Boss and Ibanez into the nineties.
User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

bmxguitarsbmx wrote: 03 Nov 2020, 01:50 Connect R105 to R106. R105 is the discharge path to ground for the detector voltage. I made this comment earlier. If you check the PCB layout, you will see that it is so.
I got it working. It actually doesn’t need a discharge resistor. I followed the Soloist schematic almost exactly but used a 0.22UF cap on the compressor. I believe the LED draws power off the Cap and that’s how it discharges? I’m not sure but it works quite nicely. You can get it so strong that every string is the same volume lol. It’s fairly easy to bias it with 10M trim and 2n4339

User avatar
bmxguitarsbmx
Cap Cooler
Information
Posts: 510
Joined: 25 Mar 2010, 21:15
Has thanked: 700 times
Been thanked: 216 times

Post by bmxguitarsbmx »

Carvindc125 wrote: 22 Jan 2021, 16:00
bmxguitarsbmx wrote: 03 Nov 2020, 01:50 Connect R105 to R106. R105 is the discharge path to ground for the detector voltage. I made this comment earlier. If you check the PCB layout, you will see that it is so.
I breadboarded the Soloist Compressor last night and got it to fire up.
But I am having an issue. Voltage to the LED will slowly rise and never go down. It just keeps going up slowly.
Oddly the compression effect works fine but eventually I have to disconnect the power supply then turn it back on. The power will go back to 0 and start slowly climbing again. Im afraid to see how high it will go. at 2.0V I pull the plug so the LED doesn't blow.
I built the compressor as drawn below with all spec components including a 2n4339.
I assume my issue is due to no path to ground for DC as you describe.
But Where do you move R105? R105 is 560ohms Its connected to the Source/R106 already, and its helping to set the gain of the op amp.
To my eyes it looks like R111 (56K) should be moved and connected from Source of JFET to Ground? Or from R107 to ground in parallel to the cap?
What do you think? See schematic below.
Anyway any help here would be awesome! If I can fix this LED from trying to blow up I'll have a fully functional compressor!
There are several different schematics floating around on this preamp. If you followed this schematic, the discharge path to ground is R111. You could move it directly to the jfet drain, but not the source. While I have messed with this circuit in the past, I don't remember exactly how I had it set up. I have confirmed that there are several wrong schematics based on tracing the PCB in the very same .pdf that has the incorrect schematic. If you peruse the schematics, you may find variations that have your missing pieces. Otherwise, trace the PCB. leaving errors in service schematics is a common thing.

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

bmxguitarsbmx wrote: 03 Nov 2020, 01:50
There are several different schematics floating around on this preamp. If you followed this schematic, the discharge path to ground is R111. You could move it directly to the jfet drain, but not the source. While I have messed with this circuit in the past, I don't remember exactly how I had it set up. I have confirmed that there are several wrong schematics based on tracing the PCB in the very same .pdf that has the incorrect schematic. If you peruse the schematics, you may find variations that have your missing pieces. Otherwise, trace the PCB. leaving errors in service schematics is a common thing.
I figured it out.
I used this schematic of the Soloist and replaced R106 and R107 with a 10M trim Pot.
Pin 1 to Source Pin 2 to Gate Pin 3 goes to Cap/Diode.
I used the 2n4339 JFet and with the 10M pot the compressor works straight away. You can very the strength and Gain of the compressor by adjusting the trimpot. 3/4 of the Trim Pots range is perfectly usable depending on how strong you want the compression. The last 1/4 turn starts to introduce distortion and things get nasty, In my case to get the Compression to sound the same as my soloist the Trim pot is left just below the point where strumming a hard chord causes breakup.
10M trimmer and the 2n4339 is almost a fool proof system. Very easy to adjust and doesnt really go out of bias. I tried 5 different 2n4339s and though the sweet spot was different for everyone they were all usable with the 10M trim pot.
Attachments
Soloist Compressor.PNG
Soloist Compressor.PNG (36.57 KiB) Viewed 4066 times

User avatar
bmxguitarsbmx
Cap Cooler
Information
Posts: 510
Joined: 25 Mar 2010, 21:15
Has thanked: 700 times
Been thanked: 216 times

Post by bmxguitarsbmx »

That's awesome! Stoked to hear of your success, especially since you have a unit to compare to.

Once clarification: Are you using the 2n4339 backwards? Source pin should be connected to ground, right? I hear that you can use Jfets either way. Maybe reversed is best?

User avatar
enterprice
Information
Posts: 5
Joined: 23 Dec 2016, 20:53

Post by enterprice »

Has anyone built the chorus part before?

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

bmxguitarsbmx wrote: 26 Jan 2021, 21:56 That's awesome! Stoked to hear of your success, especially since you have a unit to compare to.

Once clarification: Are you using the 2n4339 backwards? Source pin should be connected to ground, right? I hear that you can use Jfets either way. Maybe reversed is best?
Correct.
the Drain is running to ground.
Also I have been working with someone who is an expert with compressors so to speak and together I think we have figured out how to bias any JFET not just the 2n4339.
I'll let you know very soon. We have to test first.

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

enterprice wrote: 30 Jan 2021, 19:44 Has anyone built the chorus part before?
I just build it on the breadboard last night. I am going to test it here shortly.

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

enterprice wrote: 30 Jan 2021, 19:44 Has anyone built the chorus part before?
There are like 13-14 power terminations. I can’t even remember.
Breadboarding it is not for the faint at heart lol
Attachments
6AD64C7C-402B-4BC9-879E-FB06E641AD2B.jpeg
6AD64C7C-402B-4BC9-879E-FB06E641AD2B.jpeg (27.79 KiB) Viewed 3863 times

User avatar
enterprice
Information
Posts: 5
Joined: 23 Dec 2016, 20:53

Post by enterprice »

I also have the chorus on the Breadboard. It worked in one go, only after 1 minute of playing the sound disappears. Only when I switch off and on the power can I play for another minute and that keeps repeating. I just can't figure out what's causing it.

User avatar
Yazoo55
Breadboard Brother
Information
Posts: 130
Joined: 27 Apr 2008, 11:35
Location: Morpeth, Northumberland, U.K.
Has thanked: 11 times
Been thanked: 27 times

Post by Yazoo55 »

I built the chorus board and the problem I had initially was a faulty MN3101 which got really hot. The sound was really distorted. Fortunately the BBD survived and it all worked when I replaced the MN3101.

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

Yazoo55 wrote: 05 Feb 2021, 11:25 I built the chorus board and the problem I had initially was a faulty MN3101 which got really hot. The sound was really distorted. Fortunately the BBD survived and it all worked when I replaced the MN3101.
There is no mn3101 in the Rockman chorus. Just a mn3007, cd4013b and a 555 timer

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

enterprice wrote: 03 Feb 2021, 12:10 I also have the chorus on the Breadboard. It worked in one go, only after 1 minute of playing the sound disappears. Only when I switch off and on the power can I play for another minute and that keeps repeating. I just can't figure out what's causing it.
Mine was whistling.
Guitar signal lost after input of mn3007.
May possibly gotten fakes.
I spent all day today rebuilding it from the ground up with all new chips I’ll try it out later. I need a break! lol

User avatar
Yazoo55
Breadboard Brother
Information
Posts: 130
Joined: 27 Apr 2008, 11:35
Location: Morpeth, Northumberland, U.K.
Has thanked: 11 times
Been thanked: 27 times

Post by Yazoo55 »

I meant the echo/ chorus board of the Rockman X100.

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

Yazoo55 wrote: 07 Feb 2021, 02:22 I meant the echo/ chorus board of the Rockman X100.
Oddly enough I’m burning up mn3007 chips. I don’t know why.
Chorus works for about 5 seconds then the mn3007 will start to heat up then the signal dies and here comes the smoke! It’s just 6 volts.
That chip is supposed to handle up to 15v so I’m wondering if I got fake mn3007s

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

Yazoo55 wrote: 07 Feb 2021, 02:22 I meant the echo/ chorus board of the Rockman X100.
On the chorus I noticed the is no ground references for the 7555, cd4013b, and mn3007
Is there supposed to be? Is there anything you did different from the schematic to get it working?

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

Does this look right according to the schematic? This is how I have the mn3007 on the breadboard but they start to burn up after 5-10 seconds
Attachments
7753F534-0D7D-45A9-ABB2-924628AD4727.jpeg
7753F534-0D7D-45A9-ABB2-924628AD4727.jpeg (119.64 KiB) Viewed 3589 times

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

In regards to the Compressor.
The big secret is the JFETS were pre selected at the factory.
We have done some extensive studies with the compressor and have learned several things
1. The R106-R107 voltage divider is for cutting distortion not biasing.
2. To get the compressor to work with minimal effort you need to choose a JFET with a VGSoff of no less than 0.3V and no more than 1.2V
3. R107 can almost always stay the same value as depicted on the schematic. R106 will need to be adjected so that at your Hardest chord strum you barely get the LED (D114) to light so that would be 1.8-2.1V on the LED at Hardest strum. That will provide you the most clean tone you can get. R106 will usually be somewhere around 1M to 5M in my experience. You can use a 5M POT to dial in the value you need then replace the pot with the respective resistor or leave the pot on as a "Dirt/Boost" knob.

So how do we get the VGSoff of a JFET? there are two main ways as far as im concerned.
1. Quick way that takes about 5 seconds is get a DCA75 Pro tool plug in the JFET and it will give you the VGSoff and more.
2. The old school manual way (takes 45 minutes to an hour) this is how I did it the first couple times after that I got the tool! But I will provide you instructions on how to do this manually.
After you use this method a few times you will come to know why Tom always used 2n4339 FETS. All of them have VGSoff between 0.6 to 1.2V which makes them plug and play in this circuit. But suppose you don't want to spend $12.00 on a single JFET.
Well get your old J201s out or what ever you have. Below is how to manually find VGSoff

How to manually get the VGSoff of a JFET
First you will need to make a circuit on a breadboard or perfboard like below. Then at +V apply +6V and -V apply -6V...look familiar? Yup! Thats the rockman power supply! It doesnt matter if you wire lug 1 or 3 to ground but pay attention to which way the pot needs to be turned in order to increase/decrease voltage.
Now comes the HORRIBLE I mean FUN! part. Use a multi meter and place the Red probe on test point A and the black probe to ground very carefully turn the pot until you get 0.3V on A
Now move the red Probe to test point B You should see anywhere from 0.01 to 0.1V on B
Go back to A and turn the pot to 0.4V then check B again. You will go back and fourth between A and B doing this and noting every change on A until you see B suddenly start to spike. If B does not start to spike by the time you get to 1.0V then toss it. That JFET will not work in the Rockman Compressor.
Say it does start spiking around 0.5V Say A=0.6V and B suddenly went from 0.1 to 0.3V keep measuring. and noting until suddenly you get a sharp spike example you get little change from B until you hit 1V then suddenly B goes from 0.4V to 0.9V or the two test points match. As long as the big spike or matching happens before 1.2V MAX then you got a winner!
Now what? Remember we noted all the readings of A and B Now put those in an Excel sheet and chart them
Take a look at the chart of a 2n4547 JFET. I tested 5 of them in this way and found 1 that had a good VGS range for the rockman. Looking at the chart we can see that the VGSoff for that JFET is right at around 1.1V that's very close to our MAX range but it will work. Notice we start to see B change Voltage once A 1.05V and then there is a SHARP increase at 1.1.
The reason we chart in Excel is to see the graph and also to precisely pin point VGSoff.
Here are my actual readings of that JFET.
2n5457 Fet#2
Point AV
0.75
1.00
1.04
1.07
1.08
1.09
1.12
1.13
1.14
1.15

Point Bv
0
0.01
0.02
0.03
0.04
0.04
0.13
0.39
1.00
1.42


Now that we know the VGSoff is below 1.2V we can plug this JFET into the circuit and then adjust R106 as mentioned above.
See chart and test schematic for references. I hope this helps someone out with this compressor.
Attachments
JFET_Test_SCHEM.png
VGS_Off_Graph.png
VGS_Off_Graph.png (13.72 KiB) Viewed 3556 times

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

I crossed referenced the JFET I graphed above with the DCA75 as well to ensure I had correctly got the VGSoff of the FET and the tool returned VGSoff at 1.13V so noting your voltages at 0.01 increments at Test point A and charting A and B in Excel will give you an incredibly accurate VGSoff voltage. Albeit at the cost of about 45 minutes and some patience.

User avatar
Manfred
Tube Twister
Information
Posts: 1936
Joined: 04 Apr 2009, 23:42
Has thanked: 1668 times
Been thanked: 1344 times

Post by Manfred »

Q101 (2n4339) acts as a voltage controlled resistor and determines the momentary gain of the compressor stage depending on the gate voltage.
The feedback loop with R106 and R107 linearizes and extends the resistance range of Rds.
The ratio R106/R107 can be used to make the slope of the linearized resistance curve steeper or flatter.
I think that for another JFETs , R109 and R105 must be adapted to achieve the same compression ratio.
For this it would be necessary to have the measured compression curve of the original X100.
2N4339 Rds versus -Ugs.png
2N4339 Rds versus -Ugs.png (15.81 KiB) Viewed 3500 times

User avatar
Carvindc125
Breadboard Brother
Information
Posts: 63
Joined: 08 Mar 2015, 21:22
Been thanked: 23 times

Post by Carvindc125 »

Manfred wrote: 09 Feb 2021, 00:04 Q101 (2n4339) acts as a voltage controlled resistor and determines the momentary gain of the compressor stage depending on the gate voltage.
The feedback loop with R106 and R107 linearizes and extends the resistance range of Rds.
The ratio R106/R107 can be used to make the slope of the linearized resistance curve steeper or flatter.
I think that for another JFETs , R109 and R105 must be adapted to achieve the same compression ratio.
For this it would be necessary to have the measured compression curve of the original X100.
2N4339 Rds versus -Ugs.png
You can use any Fet that is at or below 1.2v vgsOff.
Here is the long version of how it works.
Per Vivek my resident Rockman compressor expert
“ Design note on compressor in Rockman, with focus on linearisation of the FET; and how to set R106 and R107

A lot of folk over the years have said they could not get any other FET to work in the Rockman compressor besides the original

And could not bias the particular one that they had in their hands

After reading various technical papers from manufacturers (which are dated around 1975, so Tom most probably had access to them), I feel as follows:

The resistance of a FET Rds depends upon 2 things

A) the voltage on the gate Vgs, duh !
B) the drain to source voltage Vds

In the Rockman circuit, you can see that the Vds is the audio signal

So if we keep Vgs fixed, the Rds will still vary with the input signal. This will create distortion.

we need to do something to increase linearity and reduce distortion

It's normal to use feedback to control distortion

So we place R106 from source to gate as a feedback resistor to pass some of the source signal to the gate.

The literature says that the first guess is that it should be same value as R107

But now R106 and R107 make a voltage divider as far as the control signal goes.

Hence after adding R106, we have to make sure that the maximum voltage reached on the gate is close to the pinchoff value of the FET.

But our divider divides by half, so the maximum voltage on the integration capacitor should be about double the Vp pinchoff.

We need to place an LED D114 (Or string of LED and diodes) as a "voltage clamp"so that max voltage on integration cap = 2 times the pinchoff voltage of the FET.

Hence if we have a FET that pinches of at 0.9V, a red LED will be twice that when about 5ma is going through (About 2.1V)

But if we have a FET that pinches off at let's say 1.2V, we need about 2.4 V max at the cap, so we add one Si diode in series with the red LED.

FET have lot of variability in their specifications. Even different individuals with same part number can be quite different.

(I feel that most problems people had was not enough control voltage for that particular FET in their hands)

BUT WAIT, the schematics say that R106 is chosen in factory to suit the particular FET on that board.

The literature says that the initial guess is R106 = R107 and that is surely Ok for smaller audio input.

But to lower the distortion further for larger signals, we have to optimise R106 for that particular FET on the board

I don't believe that R107 has to be adjusted. It just needs to be large, to not load the FET.

There were some people who had success with a trim pot instead of R106 and 107. The drawback of that is when you increase R106, you reduce 107

Hence I feel correct way is

If you choose R107 = 5.1 meg fixed

Then to adjust, R106 can be about 3 meg resistor with about 5 meg trimmer in series

And you adjust for least distortion level at highest strumming levels.

When R106 and 107 are determined, we calculate the value of the integration cap such that (R106 + R107) x C106 = about 2.5 seconds.

Last part : some circuits have switches that bypass some of the Si diodes in the diode chain, so the cap voltage will always be lower than what's needed for complete pinchoff. Hence the compressor won't compress so well the loudest notes of the input. It will allow loud notes to go through. Hence we have different compressor response if we short out some of the diodes in the diode change

NB: all above subject to change when someone corrects me with better vision on how to calculate the components of the Rockman compressor.”
The diode/LED pair along with the in the Soloist/X100 are what set the 1.2V If you want to go above that then you will need to change the LED/Diode arrangement by adding a diodess in series until you reach the proper clamp voltage for the vgsOff so if you have a vgsOff of 1.9 then your clamp needs to be set at 3.8V
The schematic has it set at 2.1 so we need to up it by 1.7 volts so you would add an LED or combo of diode/led in series to get to 3.8V clamp

Post Reply