auto radio solid state vibrator for vintage tube radio

[PM3295]

Fresh mind always works. No rush.

[sphinx]

i think this is as clean i am going to get it i think

because of having that ground compoents close to each other  to not cause issues or behaving weird

i also placed a couple of compoents on the bottom side to make it cleaner routingwise

i use thru hole transistors for easier repair design can be modified to have SMD's as well

info about the circuit is all in text please be aware of pinout on mosfet and change schematics and pcb if nescessary

i modified the R11 for better power handling (10w) smaller size can be used if wanted or can be shorted as well

board is now 2000x1200 mils

have fun with the design

[PM3295]

It looks good. On my design, I would maybe do a different placement for R11 as shown (*) in the pic. It shortens the trace between R11 and FETS, which I prefer. The increased length of the R10 trace is not a problem.

[PM3295]

After thinking more about the intended use in a car system, I think we should rather tie the top end of R8 back to the Vref pin. This way, we will not have duty cycle variations as the charging system varies the Vbat voltage. We only need about 3.6 V back to the Comp pin for full duty cycle, so using the 5.1 V ref output should work. R8 will change to about 2.7k for this. I will check it out and confirm.

Doing this mod on the pcb should not be difficult.

[sphinx]

i modified R11 on the board will wait with upload until you come back with confirmation

[PM3295]

Ok, we can leave as is. Since the variation is picked up by the feedback pin as well, the duty cycle adjusts in the right direction. With increased Vbat, the duty cycle is reduced, so that we have some correction on the final output of the transformer. The duty cycle increase with Vbat going lower.

Scope pic shows Vbat varied between 6 - 7.5 V and the responding duty cycle.

[sphinx]

ok here are the files with latest edits

r11 moved

have fun

[PM3295]

Looks even better in 3D!

[sphinx]

lets hope it fits in a can since i am not sure how big thos are

[PM3295]

Here are the dimensions on the most popular one out there. https://www.pekorf.com/RVB2_Assembly.html

They don't have such a large main capacitor. I don't think it needs to be 2200 uF. We can make it 220uF or so. The FETS can always be soldered from the opposite side and be bend over to save on the overall board height.

[sphinx]

i will make the big cap thru hole as well and mount it on the back side so that the fet's can be bent

Posted on: 2023-02-16, 01:46:16 - Automerged
nothing else has changed beside the cap has been made thruu-hole and placed on the backside so now both cap and
mosfets can be bent alongside the pcb and o also lef a bit of area where glue can be applied

now the pcb is even smaller than before and extremly close to his size

i used a 1000uf 25v cap wich can be changed to other size if wanted just bending the legs of bigger or smaller one

have fun with it

sadly i cant make it gorgeous with bent cap and fets so it would awesome in 3D

p.s. one can add mounting holes at the bottom of the card to make it easier to mount i wont upload with thiis because it can be very simple to add
i also realized i made pcb to mm just change back by presing "q"

p.p.s i just couldnt help myself i made the pcb as smaller to be easier to place in raios the size is now 1000x1900 mils a bit smaller than mcu version of the pcb

[PM3295]

At the moment, this design can only replace the asynchronous vibrators. If we want it to be used on synchronous types as well, we need to place two additional HV diodes with their cathodes connected to ground. The anodes go to each side of HV windings of the transformer.  The diodes should be rated at about 400 V/0.5A minimum.
Basically the diodes replace the extra set of contacts used for synchronous rectification.  

[sphinx]

man you are a good friend, just when i shrunk the pcb you just had to come up with something new. LOL

is it a schottky or normal diode ?

is your name perhaps Murphy ? you you bugger

[PM3295]

 Well, the first YT guy only worked with the simple type  . Just if we want it to be multi-purpose, we will need these diodes unfortunately. I think you can get sk diodes up to 600V last time I checked.

[sphinx]

now those 2 diodes are added on board as well, getting crowded/full on pcb, there is still some space to shoehorn something in.

i also added info about them in schematics as well.

tell me if if it looks ok.

[PM3295]

I think it is good for now.
One other thing; There is a sweet spot to set the P2 (duty), to get about 40% for the nominal operating voltage (no charging). Because the feedback and compensation action counter each other, you want to set it at the break-point where the duty adjusts in the correct sense. If you get it at the wrong side, you get the duty increasing with more voltage (comp dominates). Do you think we should have a note on the diagram for setting this adjustment? It may be a good idea to put a pad on the wiper trace of P2 to act as a test point. I will do more testing on 12 V and higher to see how this reacts. On a 6 V system, I can get a good response from 6 V up to 8 V, with the comp resistor increased to 2.2k and the wiper set at the 2.55V point. I will verify if this is good for the higher voltage systems also.

What do you think?

[sphinx]

i added a pad/hole for a testpoint for measuring on the wiper trace for easier measuring.

never hurts with information to make a circuit work better is my opinion

i also mean we have made a modular design of the circuit with lots of other info on how to use it so of course it is good to have that info as well

just tell me what tmessage i should put there on the schematics.

[PM3295]

Since the value of R2/3 will have some tolerance, and not always divide down exactly 50%, it is better to measure for a voltage difference between this point and P2 slider. It should be adjusted for +150 mV, and with R7 at 3k3 it tracks well for 6 V or 12 V once set at the nominal voltage of operation.

[sphinx]

i did the modifiations of schematics and added info as well

during the time i did some cleanup (beautying) it up.

[PM3295]

Hey man, it looks like our design work is done on this! Now we just need someone who plays with these radios to actually try it out. I don't have many valve type equipment, except for a few VTVMs, and they operate without this.

[sphinx]

i was wondering about adding some info on how it is connected to the transformers for both types on the schematic sheet

i just haave a couple of tubes somewhere in the bins but that is as far as own and valve related gadgets.

[PM3295]

If one could get hold of a 6 V vibrator type transformer, then it will be easy to make up test to see how well it generates the required HV supply. I suppose you could add a diagram to show how it needs to be connected, but I guess the people restoring these radios already know how to connect it. There are many online forums dealing with vintage radio restorations.

[sphinx]

i added info on how to connect it, better to be clear than not

this is only change to schematic sheet nothing else changes just added connection info thats all

even if people know never hurts with this info so no need to ask a ? when info is provided.

[PM3295]

And of course, if you want something simple and don't care about all the extra features of the 3525, you can modify the 4047 circuit to give you limited duty cycle control as shown.

[sphinx]

shall we create a schematic/pcb for variant ?