OK ... so you guys seem to have some experience here I don't know about. I just plugged the gerber file into https://jlcpcb.com and it cam back with a price of $8.70 plus $17 dollars shipping for 5 prototypes. Can this be correct? If so whay haven't i been making more stuff!!! .... that said ... is this thing ready to go, or is there more to change/do?
@coacharnold,
Once you start ordering pcb's, it won't be long till you attend your first Board Hoarders Anonymous meeting!
Once you start ordering pcb's, it won't be long till you attend your first Board Hoarders Anonymous meeting!
Depends on the shipping option
Fast service like DHL is the most expensive
Standard postal parcel (delivery time about 2 weeks) is definitely the cheapest
For me (delivery to Europe) PCB 7.47Euro Delivery -cheap czaczka -6.75 Euro
It's not expensive ....
Fast service like DHL is the most expensive
Standard postal parcel (delivery time about 2 weeks) is definitely the cheapest
For me (delivery to Europe) PCB 7.47Euro Delivery -cheap czaczka -6.75 Euro
It's not expensive ....
Very cool ..... so just to confirm. The Gerber file on the first post of this discussion is updated to the be the current/final design itteration?
@coacharnold
The first post#1 has all the current files.
But Zoltan will update the Gerbers PCB with a small adjustment.
The first post#1 has all the current files.
But Zoltan will update the Gerbers PCB with a small adjustment.
I wanted to add 10*4mm tht capacitors
Unfortunately, in order to fit them close to the op-amp, major changes to the PCB are required
I really can't do it today
However, I don't fully understand the problem - the datasheet recommends ceramic smd..
Why are we trying to be smarter than the engineers from BB?
Unfortunately, in order to fit them close to the op-amp, major changes to the PCB are required
I really can't do it today
However, I don't fully understand the problem - the datasheet recommends ceramic smd..
Why are we trying to be smarter than the engineers from BB?
SMD package for opamp bypass caps is the best option. 😉
For low cost boards from JLCPCB I use the global saver option, and they are usually here in less than 2 weeks. (shipping < $2 and quite reliable to the US)
I definitely belong to board hoarders anonymous. LOL. I have done almost a hundred PCB designs in the past half decade - they are so inexpensive I go straight to PCB rather than breadboard in most cases. (I like doing PCB layout, and find it entertaining as added motivation. I use KiCad)
Lineup, nice design. 😀
I'm sure I'm gonna open a can of works here, but I'm trying to get Eagle to work on Fedora 40 ... having some issues. DOes anyone have experience with Eagle PCB on linux. I had it working years ago, but cant seem to get it going?
If you don't ultimately succeed, KiCad supports several of the common Linux distros natively as well as windows and mac. I very highly recommend it.
I hesitate to post because I'm so late in coming to this thread but I feel compelled, since PCB procurement seems imminent.
I really like the approach of using an opamp as input and driver to the output stage. I think that the chosen opamp is an excellent choice. https://www.ti.com/lit/ds/symlink/opa1655.pdf?ts=1725825767901
But I believe the design, as shown, is almost certain to oscillate.
Note from the data sheet that the opamp has about 53MHz bandwidth and is unity-gain stable. Referring to the schematic below, if feedback (what's termed "Out" in OP schematic) were taken from the opamp output pin, the network would be stable and would roll off at 200kHz, as described in OP. The dilemma is that the cascade of the opamp and the output stage is not unity-gain stable, as the output stage contributes lots of added phase shift, starting at a few hundred kHz.
I believe the design can be made stable by making C2 = 0 and adding a small cap from opamp output to inverting input. I can also recommend a few other minor suggestions re the circuit and layout. I do think the layout already looks pretty good.
Similar dilemas plague the Chippewa design, which I also admire. The remedy is similar, but different since the selected opamp is decompensated and needs minimum gain of 5.
I'm sorry my comments come so late.
Best to all.
I really like the approach of using an opamp as input and driver to the output stage. I think that the chosen opamp is an excellent choice. https://www.ti.com/lit/ds/symlink/opa1655.pdf?ts=1725825767901
But I believe the design, as shown, is almost certain to oscillate.
Note from the data sheet that the opamp has about 53MHz bandwidth and is unity-gain stable. Referring to the schematic below, if feedback (what's termed "Out" in OP schematic) were taken from the opamp output pin, the network would be stable and would roll off at 200kHz, as described in OP. The dilemma is that the cascade of the opamp and the output stage is not unity-gain stable, as the output stage contributes lots of added phase shift, starting at a few hundred kHz.
I believe the design can be made stable by making C2 = 0 and adding a small cap from opamp output to inverting input. I can also recommend a few other minor suggestions re the circuit and layout. I do think the layout already looks pretty good.
Similar dilemas plague the Chippewa design, which I also admire. The remedy is similar, but different since the selected opamp is decompensated and needs minimum gain of 5.
I'm sorry my comments come so late.
Best to all.
Last edited:
So you say we remove C2 and put a cap from output to inv instead.
Are you sure?
Because my simulation works well and is stable.
Are you sure?
Because my simulation works well and is stable.
I concur BSST. Your reservoir caps are in parallel. If it is a single board, then you can drop two of them, or if you wish and it is essential for motorboating, then double the capacitance. I don't think it is necessary for all those parallel caps on the input, just select one closest to what is needed to pass the desired frequency. There will be an input off-set because your input and FB resistors are not the same value.
Last edited:
You really should do loopgain plots as well. Phase margin of the original schematic is only 23 degrees. As @BSST posted adding a small cap (e.g. 10pF) from opamp output to inverting input and removing 68pF (C3 in original) results in increased phase margin (about 63 degrees).
Simulated great amp ideas may very well oscillate in the real world, so I am eager for the first tried and tested amp report...