News:

Forum may be experiencing issues.

Main Menu

EMS synthi Hi-Fli repair

Started by bsoncini, February 03, 2019, 07:35:45 PM

Previous topic - Next topic

bsoncini

Vero layout of this thing? You are nuts

PMowdes2

Ok, got all the schematics done except the controls board.  Time to start getting stuff checked over.
DeadEndFX

JustALittleSolderBoy

Quote from: bsoncini on February 13, 2019, 10:54:19 PM
Vero layout of this thing? You are nuts

Yepp - vero! Maybe I am nuts, but I have no knowledge in designing proper pcb boards and in tools like eagle and such. So vero is the only way to go for me - unless there is a pcb available. My etching pcb experience is " too basic".

At the moment the layout of my phaser board is 52 rows x 94 holes. There are different versions of the fuzz board - the smallest is 52 rows x 76 holes.
The pedal board is 9 rows x 21 holes and so it should fit into every Crybaby-like wah shell.
The numbers of rows of both the larger boards is equal, because I intend to mount them like a sandwich. I don't plan to make a seperate wiring board for the pots and switches - parts of these are on the fuzz and phaser boards and some parts will be soldered to the pins of the mounted switches and pots.

Some parts are obsolete and will be hard to get: the N558V/NTE778A should be a standard 1458 dual opamps. EMS themselves used LM1458 in later versions of the Hifli. I think about using NE5532 instead.

But other parts are more tricky:
C746 (A6, A16 fuzz board) are almost inpossible to get - I think about LM194H/LM394H metal can devices here. I thought about matching my own FETs with a PEAK DCA70 and make some clamps from metal sheet myself...
For TIS70 (A17 fuzz board) maybe a BFS21A is worth a try?
What whould You use for 7474 (A9 fuzzboard)? Is it a HD74LS74A?
The  2N5163 FETs will be hard to get, but there a lot of possible candidates to try -  5458, 2N3819, 5958, BF245A(GSD), MPF-102(DSG), 2N3822, KP303I, BFW11,
BF245A (GSD), 5458 (GSD). So pinout may be changed on the boards.

There are some points in the service manual where I am not certain due to the poor quality of the manual. I refer to the pdf-file from http://jbemond.free.fr/SDIY/EMS/EMS_Hifli.pdf:

1) The center pole of the SW3 modulation switch (SP3T) connects to #20 of the wiring board schematic. From there it should go to the unreadable # on the phase card schematic (lower right corner of the schematic in the pdf). From there it goes to R23 2,7K. But where from there? +12V or -12V?  The handwriting there is unreadable...

2) In the fuzz card schematic there is a R108 to LED1 #8 (near lower left corner of the schematic in the pdf). The resistance of R108 - 3,3K?

3) In the fuzz card schematic there is a R105 below D15 near R108, R106 & R107 (near lower left corner of the schematic in the pdf - connects to transient vibrato too). The resistance of R105 - 27K?

4) There are pictures of the phase board where a wire goes from #20 of the phase card some where around A2 (Dual Opamp) on the phase board. Is it really A2 and where exactly leads the wire to and is there another resistor (resistance?) added?

5) There are pictures of the fuzz board where a wire jumps from #13 to A10 (3080) on the part side of the board. Where does the trace near the 3080 lead to?

6) In the schematic of the fuzz card (the upper right side) it is quite hard to read the traces around #22, 21, 19, 18, 20 and just below the 5-6-7 opamp triangle of A5 and Q1 (BC169C) around Q10 (10uF tantal) and the 470K to +12V. It is hard to decipher, what is a connection and where traces cross each other only. Maybe someone has a readable schematic or high quality pictures of the solder and part sides of the boards?

Maybe some of these questions may show my stupidity in Your eye, but as I told You before - I am a total newbie when it comes to electronics.

Any hints or comments, folks?

PMowdes2

Quote from: JustALittleSolderBoy on February 14, 2019, 08:43:35 AM
Quote from: bsoncini on February 13, 2019, 10:54:19 PM
Vero layout of this thing? You are nuts

Yepp - vero! Maybe I am nuts, but I have no knowledge in designing proper pcb boards and in tools like eagle and such. So vero is the only way to go for me - unless there is a pcb available. My etching pcb experience is " too basic".

At the moment the layout of my phaser board is 52 rows x 94 holes. There are different versions of the fuzz board - the smallest is 52 rows x 76 holes.
The pedal board is 9 rows x 21 holes and so it should fit into every Crybaby-like wah shell.
The numbers of rows of both the larger boards is equal, because I intend to mount them like a sandwich. I don't plan to make a seperate wiring board for the pots and switches - parts of these are on the fuzz and phaser boards and some parts will be soldered to the pins of the mounted switches and pots.

Some parts are obsolete and will be hard to get: the N558V/NTE778A should be a standard 1458 dual opamps. EMS themselves used LM1458 in later versions of the Hifli. I think about using NE5532 instead.

But other parts are more tricky:
C746 (A6, A16 fuzz board) are almost inpossible to get - I think about LM194H/LM394H metal can devices here. I thought about matching my own FETs with a PEAK DCA70 and make some clamps from metal sheet myself...
For TIS70 (A17 fuzz board) maybe a BFS21A is worth a try?
What whould You use for 7474 (A9 fuzzboard)? Is it a HD74LS74A?
The  2N5163 FETs will be hard to get, but there a lot of possible candidates to try -  5458, 2N3819, 5958, BF245A(GSD), MPF-102(DSG), 2N3822, KP303I, BFW11,
BF245A (GSD), 5458 (GSD). So pinout may be changed on the boards.

There are some points in the service manual where I am not certain due to the poor quality of the manual. I refer to the pdf-file from http://jbemond.free.fr/SDIY/EMS/EMS_Hifli.pdf:

1) The center pole of the SW3 modulation switch (SP3T) connects to #20 of the wiring board schematic. From there it should go to the unreadable # on the phase card schematic (lower right corner of the schematic in the pdf). From there it goes to R23 2,7K. But where from there? +12V or -12V?  The handwriting there is unreadable...

2) In the fuzz card schematic there is a R108 to LED1 #8 (near lower left corner of the schematic in the pdf). The resistance of R108 - 3,3K?

3) In the fuzz card schematic there is a R105 below D15 near R108, R106 & R107 (near lower left corner of the schematic in the pdf - connects to transient vibrato too). The resistance of R105 - 27K?

4) There are pictures of the phase board where a wire goes from #20 of the phase card some where around A2 (Dual Opamp) on the phase board. Is it really A2 and where exactly leads the wire to and is there another resistor (resistance?) added?

5) There are pictures of the fuzz board where a wire jumps from #13 to A10 (3080) on the part side of the board. Where does the trace near the 3080 lead to?

6) In the schematic of the fuzz card (the upper right side) it is quite hard to read the traces around #22, 21, 19, 18, 20 and just below the 5-6-7 opamp triangle of A5 and Q1 (BC169C) around Q10 (10uF tantal) and the 470K to +12V. It is hard to decipher, what is a connection and where traces cross each other only. Maybe someone has a readable schematic or high quality pictures of the solder and part sides of the boards?

Maybe some of these questions may show my stupidity in Your eye, but as I told You before - I am a total newbie when it comes to electronics.

Any hints or comments, folks?

We got this bro, we've generated cleaned up versions of the schematics in the service manual based on Brent's photo's.  And we are in the process of putting them into Eagle so we can make some pcb's.  Once we have everything checked over maybe we can answer your questions.  You are mad trying to do this on vero though :)
DeadEndFX

JustALittleSolderBoy

Quote from: PMowdes2 on February 14, 2019, 11:24:02 AM
Quote from: JustALittleSolderBoy on February 14, 2019, 08:43:35 AM
Quote from: bsoncini on February 13, 2019, 10:54:19 PM
Vero layout of this thing? You are nuts

...blubbdidubb...


We got this bro, we've generated cleaned up versions of the schematics in the service manual based on Brent's photo's.  And we are in the process of putting them into Eagle so we can make some pcb's.  Once we have everything checked over maybe we can answer your questions.  You are mad trying to do this on vero though :)

Thanks Phil! Maybe YOU can help me find the answers I need!
As soon Your design of the PCB is finished, then I am interested in two of these boards of course (a "DoubleFli" would be gosh, wouldn't it!?). 

But a vero has its advantages, too - it would be much easier to mod it. Here are some mods I think about:

-possible remote-ports for most the tapers and switches (ED Tap-Tempo; remote pedals)
-send-return-jacks
-listen-in-switch into the stages of the phaser section - maybe more phase stages on a seperate daughter-board can be added!?
-signal injection/ejection/connection points to add tap-tempo, sequencers; maybe signal-filters/splitters, maybe VCFCLFO, VCADSR (ED-Druid), SFH. The possibilities are endless. Ghielmetti like patch-matrix with buffers like for the AKS would be cool to split and mix signals.

I think this baby is a gold mine of sounds never heard before and there is much more beyond to be discovered...

I may share some files and some snapshots of the progress of my layouts if You like. I am not that great web user and never posted files nor pictures here before. Is it possible to upload files here or is there a reliable service (that doesn't give away everything to Google Ltd. & Co)? Or may somebody my upload on my behalf? Any help will please...

somnif

Quote from: JustALittleSolderBoy on February 14, 2019, 08:43:35 AM

But other parts are more tricky:
C746 (A6, A16 fuzz board) are almost inpossible to get - I think about LM194H/LM394H metal can devices here. I thought about matching my own FETs with a PEAK DCA70

The 2C746 is a matched pair of BJTs, not FETs. You can still find SOME matched dual BJTs on Mouser, but they are ludicrously expensive. Probably easier to just play the old "I'm building an octave fuzz" game and hand matching some. They do lave a very low hfe, which could be tricky, but I'd honestly try something like 2n3904s or 2n2222s.

Could also, if desperate, try some SMT matched pairs, those are fairly readily available. 

JustALittleSolderBoy

Yepp, you are right! My mistake - these are matched pairs of BJT of course. It should be easy to get matched pairs with a tester like Peak DCA. I ment TIS70 and not C746...

About the low hfe: I read somewhere some comments on some russian Germanium transistors, that have leak resistors built-in to ged rid of leakage. Someone hinted at the fact, that adding such resistors would lower the hfe of the transistor - thus all of these Russian babies concerned have fairly low hfe. Maybe we might be able to lower the hfe of these BJT to the level needed by adding such resistors?

And maybe TIS70 can be replaced by DIY matched pairs, too?

somnif

Quote from: JustALittleSolderBoy on February 15, 2019, 07:11:04 AM
And maybe TIS70 can be replaced by DIY matched pairs, too?

FETs are a bit voodoo in terms of what characteristics matter about them in audio situations (to me, anyway, I'm a biochemist not an engineer). But in terms of matching, its interesting.

For Phasers, the quick and easy method is to match the Vgs(off), and you typically want to get them to within ~10mV. The TIS70's FETs, by its datasheet, are matched within 32mV, so that would be easy enough with our current methods.

If you use the "Greatly improved JFET matcher" circuit and can measure Idss, then you'd look for an 80% or better similarity (Idss1/Idss2 >= 0.8). Again, not terribly difficult, though finding a pair where both these parameters are met can start to get hairy.

So all told, assuming we can find JFETs that match those characteristics found in the TIS70, finding a matched pair wouldn't be impossible (I think). The 2n5457 and 2n5458 have a similar range of specs for Vgs(off), Igss, and Idss and they're still readily available in SMT format (MMBF5457/8), though I'm not sure how you would set up the thermal pairing clip for surface mount stuff (little daughter board with them on one side and warm resistor on the other? dunno).

But again, I spend my days torturing yeast, so I'm not the expert to talk to about JFET cork sniffing. I shall leave that to the gurus. I'm attaching the TIS70 data sheet if you feel like some light reading.

JustALittleSolderBoy

Bytheway - on his website www.phutney.com Derek Revell suggests to use LM394CH and 2N2916 as possible substitutes for 2C746 for is EMS Synthi clone Phutney!

JustALittleSolderBoy

Quote from: PMowdes2 on February 14, 2019, 11:24:02 AM

We got this bro, we've generated cleaned up versions of the schematics in the service manual based on Brent's photo's.  And we are in the process of putting them into Eagle so we can make some pcb's.  Once we have everything checked over maybe we can answer your questions.  You are mad trying to do this on vero though :)

Hi Phil,

only want to ask for an update - how is the progress regarding Your Hifli schematic?

Are You ready to answer my questions?

Regards!

PMowdes2

Quote from: JustALittleSolderBoy on February 25, 2019, 02:25:22 PM
Quote from: PMowdes2 on February 14, 2019, 11:24:02 AM

We got this bro, we've generated cleaned up versions of the schematics in the service manual based on Brent's photo's.  And we are in the process of putting them into Eagle so we can make some pcb's.  Once we have everything checked over maybe we can answer your questions.  You are mad trying to do this on vero though :)

Hi Phil,

only want to ask for an update - how is the progress regarding Your Hifli schematic?

Are You ready to answer my questions?

Regards!

We have finished the schematics and i'm in the process of laying out the pcb's.  It's slow going not least because I have no idea what size enclosure I can get it into.

Drop me a PM and i'll see if I can answer some of your questions

Phil
DeadEndFX

JustALittleSolderBoy

Hi Phil,

that is good news! I will have a look into my HiFli files again and gather all the questions I have. I will send You a PM as soon I have everything.

Regards!

PMowdes2


Here's a progess update

The top board which is the psu and phaser



One of the two identical diode ladders



And the pedal preamp



Just the bottom fuzz / vibrato board to go
DeadEndFX

skyled

Please tell me that the diode ladder won't use smd diodes. It sure looks like smd in that picture.

PMowdes2

Quote from: skyled on March 09, 2019, 08:54:05 AM
Please tell me that the diode ladder won't use smd diodes. It sure looks like smd in that picture.

Afraid so, only way to keep the size down to something less than a football pitch.  We are thinking of getting these fabricated assembled, it will add some additional cost but a whole lot less headache
DeadEndFX