TM 808 White's

[JC1]

big ohms

Hi Guys,'

By the way after you get through building

the circuit clean the capacitor body and board

with alcohol to remove grime and flux.

On the layout you should try and move other

voltages from around this area, as guard rings

will reduce surface leakage currents, but the

insulation of the pcb material may matter if close.

forget the bias current on 3140 but it is low

as example 1 volt / 100 picoAmps would be

10 GigaOhms or 10,000,000,000 ohms.

your meter doesn't read this.

supply voltages of the 3140 is usually the closest

voltages.

good luck, and good catch, or you could go to

alot of trouble and end up with something worst

than orginal.

after construction check this part of circuit for drift.

[JC1]

78S40

Hi Leto,

Geo will get this but believe that is

78S40 and the 14584-B is motorola

part number and 4584 should be close

but haven't checked for each parameter.

sometimes brand matters.

[Leto]

fog clearing

Thanks JC1:
I noticed guard rings and wondered about that fatboy cap - you cleared the things perfectly...
maybe I'll move some tracks - but just a bit as I want to stay close to original..
Also: yes 78S40 fits well into shematic - I misread the part No.

[JC1]

select

Hi Leto,

Glad that helped. I looked at the schematic

and it appears alot of parts are "selected".

the 4016 hooked up to the cap and 3140 is

probably selected for off state leakage.

The 324s are probably offset selected.

You can add some offset circuits to null some

of these things since you are looking at

building this yourself.

Yes R1 should be a 22 Meg also, i believe.

staying close to orginal should be fine but

if you can make things a bit better that

could nice as well.

good luck.

[ivconic]

i am late...


I was absent few days...JC1 already gave answers, thanks JC1!

Here is post i prepared few days ago...

For now we have to stick to Estebans results,since i kept only pcb from TM 808.The rest
gone with other machine. Friend took it with him on journey. He will be back in a few
weeks,than i'll do the measurements.
"why is that c46 so fat?"
See explanation on "So whats the story with Troy's Shadow" thread, simillar reasons.
78S40 is switching regulator IC. Same stage with it,you may find on some other White's
machines like 600DI, Eagle II etc...
It is obvious that you can not supply this device only with 4.8 v or 6v batt.
78S40 stage is providing proper voltages for device, see schematic.
14584-B...yes it is same. Prefix "1.." is most often used by Motorola.Why?
I do not have clue!
I have checked about material, yes it's gonna be easy to obtain all....except maybe
quartz. Anyway i intend to try with 32kHz, which is easy to find. Look at any old
pc motherboard(286,386,486...)...clock crystal usually.
78S40 stage have one toroid also. L1=100-105 windings, 0,4-0,5mm wire...
Toroid ring 17mm outer diammeter, 9mm inner,9mm high...
I counted and measured this L1 without reassembling it,so minor mistake is possible.
L1 should be arround 830mikroH.
I made simillar voltage convertors many times.Usually when need doubler or trippler
with descent power consumption.Than you need to drive external NPN or PNP transistor
also...like in this project...

I am finishing some device this days, when finishing, i'll be back here....
best regards!
P.S.
Leto can you put both sides in one cdr and post here? Anyway thank you very much on help!

[Geo]

Hi. Ivconic you must measure the inductance of coils , so to determine the type of core. You must find core with same characteristics (mo). If it not the same you would not have good adaptation and the output and general sensitivity will be poor.
Regards

[Leto]

selected IC's ?

Somehow I hoped that these selected IC's don't matter much.
My logic (as non EE) was that differences between IC's of same batch can not be large enough to matter. ?? So probably I'm wrong ?

Ivconic - I will post all the files here on forum just need my job computer to extract the files
(You don't want to know who really payed hours working on these files he,he :o )

[JC1]

select

Somehow I hoped that these selected IC's don't matter much.
My logic (as non EE) was that differences between IC's of same batch can not be large enough to matter. ?? So probably I'm wrong ?

Yes, you are probably wrong, since they went to the

trouble to put that on the schematic. But hey don't loose

hope, put it together, turn it on and see. If you are getting

too much drift or offset, put in a pot and series resistor

on wiper and sum in some adjustment. Now if it starts out

large then will probably change more with temperature

then if it is smaller. but again a test point and field trim

it on the hot days so to speak. Give it a try it may be fine.

[ivconic]

...


"Somehow I hoped that these selected IC's don't matter much.
My logic (as non EE) was that differences between IC's of same batch can not be large enough to matter. ?? So probably I'm wrong ?"

I am not sure that understood what are you wrong about?
But if you think on IC's same type but different manufacturers, than you are
not wrong at all.
If you have doubt about mentioned prexix "1.." on Motorola's is's than dont
worry. Same as without prefix. In most cases,let's say 95% ic's are same.
TM808 list of ic's has neither one "special" IC. All of them very ordinary.
But JC1 is right when talking about possible drifts due remain original design
of pcb. Although there are oder ways to avoid drifts...
Geo i will measure all in few weeks,when my friend give me back coils...
regards

[Leto]

TM808's PCB

Finaly here it is.
All the work was done in AutoCad and exported into CorelDraw and therein converted to raster files. Attached in .zip file are 600dpi .bmp and .cdr files.

[Leto]

TOP

Preview 1

[Leto]

BOT

preview 2

[ivconic]

Bravo!


Bravo! Thanks!
Regards!

[GTK]

There will be small problem with components placement since there isn't a picture with the boards +components values

[Leto]

Parts

There will be small problem with components placement since there isn't a picture with the boards +components values

This is not such a major problem as pictures of the board are posted right in this thread ... but to make things even easier I extracted parts placement from scans and here it is.
GTK I suggest you assemble board looking closely at shematic (parts placement only for quick help) this way you'll learn much more of the project, also this way you will check the board for mistakes.