What's the best prototyping platform?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I use 1/16" Nema CE laminate, canvas & phenolic. I drill with #42 High speed drill, no diamond drill required since no glass filler. McMaster-Carr Cuts with a hacksaw or sabre saw. Mount to metal chassis with #6 screws with 1/4 nylon air tubing for standoffs.

I put the power stages on aluminum heat sinks with mica insulators under the transistors. See heatsinks.us or the surplus houses apex and electronicsurplus. I use 30-26 ga kynar wire for early stages and 18 ga for the collector and emitter of the output transistors. I solder point to point. Leaded parts are required.

Keeping input and output wires apart and non-parallel seems to avoid oscillation so far. Input on one end, output on the other end of the driver board. Wires to OT's on heat sink fly through space, no parallal runs and especially no flat cable used. Use 33 pf short across input to keep out RF, zobel coil on output for same reason. Metal case is grounded. I use metal file boxes or scrap motor drive enclosures. VCR cases are good.
 
Last edited:
Moderator
Joined 2011
I want to build and optimize a 100W amplifier without resorting to PCB fabs.

Lay out the board with a pcb program. Using a printout as a template, center punch each hole on the blank (no copper) pcb board material. Then drill, using as few hole sizes as possible. A 0.040" hole works well for many smaller parts. Use the component leads to wire on the bottom. Of course you could just use perf board instead, with somewhat rougher results.

You could also mount turrets in the holes instead, and wire the parts to them. 2101-3-00-80-00-00-07-0 Mill-Max Manufacturing Corp. | Connectors, Interconnects | DigiKey
 
Last edited:
Hi

I use vero-board and SMD components. When I prototyped out "My Bridge Amp", more pics HERE. I Used 0603 or 0805 components and SOT-23. For the SOIC-14, I cut the pads around the holes in halve to meet up with the pin spacing. High reliable part density can be done by prototyping this way. I constructed an entire stereo using this method HERE using parts primarily that I scraped from junk. Using solder bridges from pad to pad to make the circuit traces, it is easy to change around the circuit routing.:) Also you can re-use the solder after the prototype ends up in the parts bin to use on the next prototype circuit you make. Plus, you can use jumper wires on the top side effectively giving you multiple layers. Any general standard through hole package will fit in these prototyping vero-boards and did I mention they are cheap?:D
 
Last edited:
The problem I have is that my circuit is highly prone to parasitic oscillations.
On a normal solderless breadboard it is fine but when I attempt to make it high power the leads to the output devices need to be longer which makes it much more unstable.


Admittedly I am not the best at compensation analysis since I don't usually have a need for it.
 
A schematic and picture of your layout could allow us to provide tips on your layout.
See this point to point board: Retro Amp 50W Single Supply
The in the third picture of post 212, the base drive to the output transistors is the blue wire on the right. The temperature pickup from the diode stack over the output transistors is the yellow wire on the left. This could be a major feedback loop for oscillation, but I had the wires separated by about 2" flying through the air. Plus on the temperature sense to the driver stack I put a 7 turn inductor from a switcher power supply on the signal coming back in. The temperature sense doesn't need to change in the megahertz or even kilohertz, changes of 1/5 hertz time are adequate. Inductors have higher impedance at high frequencies than low frequencies.
You'll see the input transistor TO92 lower left, the VAS TO220 upper left, the driver transistors TO220 over on the right of the card about 2.5" away from the input. No ground plane required to prevent oscillations, this is bare Nema CE board.
I don't do compensation analysis either, or sims, I just followed dynaco's pattern on c3 the 47 pf from c to b on input transistor. Dynaco also had 50 pf c to b on later versions of the drivers to slow slew rate, I didn't have to do that. TIP41C drivers are pretty slow anyway, Ft is 3 mhz. If I upgrade to Ft 100 mhz drivers I might have to deal with the compensation to prevent oscillation.
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Sharpie as resist on copper clad phenolic board. Can be cut with utility knife and ruler and and use SMT where possible. No drilling required for SMT.

For more complex layouts use either layout program or draw program. I use PowerPoint to make pretty nice layouts for amps with up to 7 or 8 actives.

Print mirror image using laser printer to vinyl shelf liner self adhesive. It releases nicely. Then iron into hot copper, then etch with muriatic acid and hydrogen peroxide mix.

Here is example of Aksa Lender Preamp;
646265d1510907783-aksas-lender-preamp-40vpp-output-smt-carbon-fb-jpg


Here is a Class A amp - DLH:
633540d1504516574-dlh-amplifier-trilogy-plh-jlh-amps-dlh-pcb-layout-00-jpg


633542d1504516574-dlh-amplifier-trilogy-plh-jlh-amps-dlh-build-02-jpg
 
Last edited:
SRBP board and the "laser toner transfer" method. Works very well if you have a little patience. The main bugger is drilling the holes !

edit: SRBP because it's cheap, easy to cut, doesnt need carbide drill bits to drill..
 
Last edited:
Design and manufacture a universal PCB following common push-pull amplifier topologies. Make the trace similar to schematic so to be easy to follow. But I found that ordering new PCBs is more expensive than buying certain ready-made amp PCB that is sold at cheap price for a reason (one mistake in the trace, unstable design, etc).


The problem with commercial circuits is they are drawn unlike a schematic so the prototyping is not fast and a little confusing. I would want one horizontal line on the top for V+, probably with RC filters at every stage, another horizontal line at the bottom for V- and may be some space when a simple R will be replaced with ccs in daughter board.


I have been thinking of the best way how to make such PCB is usable for both TO-3 and TO-240, also GSD/GDS/BCE, common source/emitter or common drain/collector.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.