That sounds intriguing Douglas! I thought one only had a choice of 2-layer, 4-layer, 6-layer etc but now that you mention it, I can see some good points for 3-layer!
Can board houses handle such things as a matter of routine?
Jan
Not to my knowledge and pcb fabricator guideline docs. Show me the stackup drawing with materials involved.
I always provide a stackup drawing as part of the drill drawing/schedule gerber to the fab as part of the data set.
Then find a place ( link to web site) to fab it, at what costs & for what advantage? imo a waste of time & $
imo you can effectively do the same thing with 4 layers, inner 2 layers at the same potential with a very thin prepreg. I have done this for high speed digital designs. A 16 layer stack up in 1/16".
With $2 pcb these days, why make life complicated and expensive when it does not have to be.
I always provide a stackup drawing as part of the drill drawing/schedule gerber to the fab as part of the data set.
Then find a place ( link to web site) to fab it, at what costs & for what advantage? imo a waste of time & $
imo you can effectively do the same thing with 4 layers, inner 2 layers at the same potential with a very thin prepreg. I have done this for high speed digital designs. A 16 layer stack up in 1/16".
With $2 pcb these days, why make life complicated and expensive when it does not have to be.
Need to know bases I guess.
Not too much to it really, just need to understand the material stackup, padstacks and types of planes if you use them. If you are using transmission lines (controlled Z), then you need to know your trace widths, dielectric constants/ spacings. Some advanced tools have these calculators built in.
Not too much to it really, just need to understand the material stackup, padstacks and types of planes if you use them. If you are using transmission lines (controlled Z), then you need to know your trace widths, dielectric constants/ spacings. Some advanced tools have these calculators built in.
You always want a symmetrical stack-up about the centerline of the PCB thickness. This applies not only to copper layers and laminate thicknesses but also plane vs signal layers. If the stack-up is unbalanced, the board will warp as it cools down out of the press.
You basically pay for materials and time in a PCB fab house. If they need to keep the PCB in the press for cool-down, you might as well just go to a 4-layer PCB as the cost will be a wash. Then you have the advantage of a power plane and a tightly-coupled ground plane on the internal layers to take advantage of some embedded capacitance.
I have designed an 11-layer unbalanced PCB with unbalanced blind & buried via structures for an R&D project and the cost was ridiculous! It was an awful price to pay for "I told you it could be done!"
But having said all of that, 3-layers is symmetrical if the center layer resides in the center of the stack-up.
But cost would be the same as a 4-layer since the process would not change other than imaging and etching another copper layer. Whereas 2-layer is simply a core that is imaged & etched on both sides and very inexpensive to produce. Once you start laminating you get into higher scrap rates and the cost goes up accordingly.
Last edited:
In earlier editions, there is a short reference to "three layers" pcb:
one solution is to use what I call a ‘three-layer board’. The mix buses are on the bottom of the PCB, the top layer above it carries a section of ground plane, and the fader connections are made by wire links above that.
So basically two layers + links
one solution is to use what I call a ‘three-layer board’. The mix buses are on the bottom of the PCB, the top layer above it carries a section of ground plane, and the fader connections are made by wire links above that.
So basically two layers + links
Hi Douglas,
I bought your RIAA electronics book recently but haven't had time to crack it yet. By chance did you take a step sidways to pull in reproduce head electronics?
As far as recording electronics goes, you mostly have a sort of power amplifier that is isolated from the record bias by a tank circuit. The bias voltage can be quite high and commonly runs around 100 KHz.
Fostex R-8? If you can, sometime switch to the Tascam equivalent. A 1/2" format is much, much better. If you can find a Tascam 58, your quality of recorded signals would go way up. A guy like you could really improve the electronics well past the intentions of the manufacturer. The 58 provides servo controlled back and take up tensions. By comparison, the R-8 just throws tape in the general direction. An old 1/2" 80-8 is even much better
I used to do warranty service for both Teac/Tascam and Fostex, plus Revox. Man, I would love to hear what you could do with a Revox A-700 or PR-99. Even a B-77 would be cool.
It's really too bad that Nakamichi never made an open reel machine. That would have been killer good!
-Chris
I bought your RIAA electronics book recently but haven't had time to crack it yet. By chance did you take a step sidways to pull in reproduce head electronics?
As far as recording electronics goes, you mostly have a sort of power amplifier that is isolated from the record bias by a tank circuit. The bias voltage can be quite high and commonly runs around 100 KHz.
Fostex R-8? If you can, sometime switch to the Tascam equivalent. A 1/2" format is much, much better. If you can find a Tascam 58, your quality of recorded signals would go way up. A guy like you could really improve the electronics well past the intentions of the manufacturer. The 58 provides servo controlled back and take up tensions. By comparison, the R-8 just throws tape in the general direction. An old 1/2" 80-8 is even much better
I used to do warranty service for both Teac/Tascam and Fostex, plus Revox. Man, I would love to hear what you could do with a Revox A-700 or PR-99. Even a B-77 would be cool.
It's really too bad that Nakamichi never made an open reel machine. That would have been killer good!
-Chris
Thank you for your kind words!
I have no plans to look at so-called super regs (I assume we're talking ultra-low noise and ripple here) because in my experience they are wholly unnecessary.
Decent opamps, and for discrete transistor circuitry some RC filtering, does the job. Nothing to be gained.
No, its always multiples of two. A three-layer PCB is a double-side PCB with links. Not as daft as it sounds- very useful in mixer design.
Yes 3 layer is doable by using 32mil pcb and mirror two boards, one of them 'single sided'
Then stack as any other multilayer board. I suppose you could have a vertical bus too through the stack.
I just had some 32mil boards fabbed with 2oz copper as an experiment.
They are stronger than expected so sticking with for the next few designs.
Then stack as any other multilayer board. I suppose you could have a vertical bus too through the stack.
I just had some 32mil boards fabbed with 2oz copper as an experiment.
They are stronger than expected so sticking with for the next few designs.
To understand pcb stack ups, from an actual manufacturer, take a look at this page to see how they are specified.
PCB Stackups : Multilayer PCB - 0.062 Thickness | Advanced Circuits
So you are sandwiching pre-preg material & cores together to be pressed and cured. Drilling depends on the structures of the through holes/vias.
Give up on the 3 layer pcb's.
PCB Stackups : Multilayer PCB - 0.062 Thickness | Advanced Circuits
So you are sandwiching pre-preg material & cores together to be pressed and cured. Drilling depends on the structures of the through holes/vias.
Give up on the 3 layer pcb's.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.