cowanrg said:
Read through the thread.. the first post goes through the number of boards needed.
The shopping list is here:
http://www.diyaudio.com/forums/showthread.php?postid=375491#post375491
(bottom of the first page [with 40 enabled])
--
Brian
A variety of options
The way I intend to use these boards will go something like this:
One set of mostly or fully stuffed boards will be used for evaluating the 3 things adjustable on each board: crossover points "fc", damping "Q", and gain required for each amped section of my speakers.
Once the "ideal" settings or jumper points are determined, I can stuff only those parts needed into boards. Some of this will be decided by the comments I gather in the forum, some will come from the manual and some from listening at home.
So, in use, the boards only require the components in the signal path. For evaluation, a set of boards allows MANY options to be explored.
I know there is a leap of faith required for some of this. I look back to where my knowledge used to be...and look forward to the increase in understanding I'll gain from the "pain" of uncertainty that must be overcome.
I look forward to working on the prototyping section of the board. A "Linkwitz transfer" is relatively simple, a way to actively boost low frequencies where the woofer falls off. Go to:
http://www.pvconsultants.com/audio/eq/linktran.htm
One IC and several passive devices are necessary. This board will already have +/- V on it from a power supply. Check out +/- 15V supplies from MJPA. http://www.mpja.com/
Marchand Electronics www.marchandelec.com offers kits that do things these boards will do. Their manuals offer insight into what is possible that hasn't found its way into Jens' "filter two" manual as of yet. Check them out, particularly the XM-9.
I hope these things help.
I'm thankful for what Jens and tiroth have done so far.
Please let us know of the questions you have.
The way I intend to use these boards will go something like this:
One set of mostly or fully stuffed boards will be used for evaluating the 3 things adjustable on each board: crossover points "fc", damping "Q", and gain required for each amped section of my speakers.
Once the "ideal" settings or jumper points are determined, I can stuff only those parts needed into boards. Some of this will be decided by the comments I gather in the forum, some will come from the manual and some from listening at home.
So, in use, the boards only require the components in the signal path. For evaluation, a set of boards allows MANY options to be explored.
I know there is a leap of faith required for some of this. I look back to where my knowledge used to be...and look forward to the increase in understanding I'll gain from the "pain" of uncertainty that must be overcome.
I look forward to working on the prototyping section of the board. A "Linkwitz transfer" is relatively simple, a way to actively boost low frequencies where the woofer falls off. Go to:
http://www.pvconsultants.com/audio/eq/linktran.htm
One IC and several passive devices are necessary. This board will already have +/- V on it from a power supply. Check out +/- 15V supplies from MJPA. http://www.mpja.com/
Marchand Electronics www.marchandelec.com offers kits that do things these boards will do. Their manuals offer insight into what is possible that hasn't found its way into Jens' "filter two" manual as of yet. Check them out, particularly the XM-9.
I hope these things help.
I'm thankful for what Jens and tiroth have done so far.
Please let us know of the questions you have.
Boards needed for common configurations, stereo
1st/2nd order 2-way: 2
1st/2nd order 3-way: 4
3rd/4th order 2-way: 4
3rd/4th order 3-way with 2nd order BP: 6
cowanrg: you can get a good idea of the parts needed from ver. 1 of the list (thank Brian for linking), but I am going to be updating this in the next week, so I suggest waiting to order any parts.
Duck-Twacy: yes, shipping address comes from paypal.
bg40403 has made some good comments about how the boards can be used. This is exactly how I will use them.
1st/2nd order 2-way: 2
1st/2nd order 3-way: 4
3rd/4th order 2-way: 4
3rd/4th order 3-way with 2nd order BP: 6
cowanrg: you can get a good idea of the parts needed from ver. 1 of the list (thank Brian for linking), but I am going to be updating this in the next week, so I suggest waiting to order any parts.
Duck-Twacy: yes, shipping address comes from paypal.
bg40403 has made some good comments about how the boards can be used. This is exactly how I will use them.
Just 2 be sure.
I want to make a 4th order 2-way stereo x-over (with discrete opamp boards)
So I am ordering:
4 MOX boards a $7 = $28
4 sets of 8 opamp boards a $5 = $20
8 sets of 4 * 2SK389 a $10 = $80
shipping to Holland $11
total $ 139
Is this right, I'm not sure about the number of sets of 2SK389. perhaps it must be 4 sets of 4??
I want to make a 4th order 2-way stereo x-over (with discrete opamp boards)
So I am ordering:
4 MOX boards a $7 = $28
4 sets of 8 opamp boards a $5 = $20
8 sets of 4 * 2SK389 a $10 = $80
shipping to Holland $11
total $ 139
Is this right, I'm not sure about the number of sets of 2SK389. perhaps it must be 4 sets of 4??
Yes, that is exactly right. You don't really need that many opamps, although it certainly won't hurt; I personally would eliminate the redundant buffer on the bandpass board (saves 4 total--use aux inputs) and probably the output gain adjust buffers (saves 8 total--replace with jumpers). This is a savings of 12 opamps.
You may order even if you had not previously signed up in the wiki.
Please make your payments now
http://www.anidian.com/audio/construct/mox_order.shtml
I would like to have all payments made by midnight Saturday, May 8th.
Thanks to everyone who has paid so far. About 50% of the orders are in.
Please make your payments now
http://www.anidian.com/audio/construct/mox_order.shtml
I would like to have all payments made by midnight Saturday, May 8th.
Thanks to everyone who has paid so far. About 50% of the orders are in.
Tiroth,
have you resolved the gain mystery? I am also thinking of the 4.7 k resistor and about current capability.
About the 4.7k - I thought it was on the Mopsamp board, not on the filter board? In any case - would this not indeed form a voltage divider with the following stage, which would shift both Fc and dB at the same time (Fc being determined by the ratios of resistance and capacitance)?
About bias and current: Am I right that bias of the Mopsamp output stage is 6.5 mA? Would that no cap the current swing to this value, and be dangerously close to the 3.5 mA highest current demand of the filter by Jens' simulation?
In other words, how about halving R8 to double the bias of T5?
Finally, for use as an output buffer, and regarding Fred's earlier objections, is it safe to use higher power devices for T4 and T5 and bias the output at a decent value, say 100 mA, without endangering open loop performance? (for +- 15 V that would dissipate 3 W and plasic devices probably won't handle that even if they are rated at high enough peak currents, so one would need a heatsink)
Oh yeah, one more. In cascading 2 boards to get a 24 dB/oct LP or HP, what is the ground plane situation? Say, imagine the board design horizontally with signal goung from left to right in case of an input>LP>HP>output buffer connection. If you want to cascade two LP and 2 HP filters on 2 boards, you have to do it "vertically" between two boards: input board 1> LP board 1 > LP board 2 > output board 2 AND input board 2> HP board 2 > HP board 1 > output board 1. Will that mix'n'match not create havoc with the grounding? (I still had a previous version in mind with 2 HP or LP on the same board, when I started thinking about how to wire up the final version...)
Sorry I couldave thought of this earlier (before I ordered
) ...
have you resolved the gain mystery? I am also thinking of the 4.7 k resistor and about current capability.
About the 4.7k - I thought it was on the Mopsamp board, not on the filter board? In any case - would this not indeed form a voltage divider with the following stage, which would shift both Fc and dB at the same time (Fc being determined by the ratios of resistance and capacitance)?
About bias and current: Am I right that bias of the Mopsamp output stage is 6.5 mA? Would that no cap the current swing to this value, and be dangerously close to the 3.5 mA highest current demand of the filter by Jens' simulation?
In other words, how about halving R8 to double the bias of T5?
Finally, for use as an output buffer, and regarding Fred's earlier objections, is it safe to use higher power devices for T4 and T5 and bias the output at a decent value, say 100 mA, without endangering open loop performance? (for +- 15 V that would dissipate 3 W and plasic devices probably won't handle that even if they are rated at high enough peak currents, so one would need a heatsink)
Oh yeah, one more. In cascading 2 boards to get a 24 dB/oct LP or HP, what is the ground plane situation? Say, imagine the board design horizontally with signal goung from left to right in case of an input>LP>HP>output buffer connection. If you want to cascade two LP and 2 HP filters on 2 boards, you have to do it "vertically" between two boards: input board 1> LP board 1 > LP board 2 > output board 2 AND input board 2> HP board 2 > HP board 1 > output board 1. Will that mix'n'match not create havoc with the grounding? (I still had a previous version in mind with 2 HP or LP on the same board, when I started thinking about how to wire up the final version...)
Sorry I couldave thought of this earlier (before I ordered
) ...MBK said:
At the moment I am chalking it up to a problem with the assembly or component. I am building up another board to check it out.
You can place it on the opamp board or on the mox board, your choice. I put it on the mox board so I could easily plug and unplug the opamp.
I ran a simulation of with the effect of the 4k7 resistor and it had no impact on Fc. I am not very concerned about Fc shift as it is tiny (1/10 octave) at high Q (1.7).
I'll run the numbers for you myself tonight. I am not concerned about the bias, the higher the current the easier it is for a class A output stage.
This is certainly an option for the more ambitious among us
. I don't understand why you went with such a complex signal routing. Why not simply do
input board #1 > LP #1 > aux input #2 > LP #2 > Out #2
input board #1 > HP #1 > aux input #2 > HP #2 > Out #2
Input grounds all terminate in board #1, output grounds all terminate in board #2.
Duck-Twacy said:Neat ordering system
Btw please could you send a mail as soon as you are shipping (there is no hurry)
Thanks.
If you've paid for your order with Paypal and gotten to the confirmation page, I have your order.I am not sending email confirmations for PayPal payments except when people write me directly with questions. I will post here and send an email when the boards ultimately ship.
I am planning on placing the PCB order on the 10th. I have plenty of 2SK389 stock.
Some distortion plots for the MOX board. This is relevant to the opamp too--since these graphs represents the signal passing all the way through (input buffer, high pass filter, output buffer) the opamp distortion will be even lower.
The problem is that the distortion is really too low for me to measure. I've plotted
Loop back - red trace
MOX - blue trace
MOX - residual - green trace
The only distortion even measurable is in the small signal measurement, and appears to be 2nd harmonic at least -88dB down (0.004%).
The problem is that the distortion is really too low for me to measure. I've plotted
Loop back - red trace
MOX - blue trace
MOX - residual - green trace
The only distortion even measurable is in the small signal measurement, and appears to be 2nd harmonic at least -88dB down (0.004%).
Attachments
Here is the measurement at 1.8 Vrms. Even though no distortion is even visible once the residual is subtracted, we can conclude just from inspection that the 2nd harmonic must be at least -87dB down, the 3rd -90dB, the 4th -106dB, the 5th -104dB (and probably much lower). That means the overall distortion is no higher than 0.0055% at 1.8Vrms, and is very likely considerably lower.
I've posted large versions of both graphs if others would like to take a closer look.
http://www.anidian.com/imgs/mox_dist_small_blowup.GIF
http://www.anidian.com/imgs/mox_dist_large_blowup.GIF
I've posted large versions of both graphs if others would like to take a closer look.
http://www.anidian.com/imgs/mox_dist_small_blowup.GIF
http://www.anidian.com/imgs/mox_dist_large_blowup.GIF
Attachments
One last post tonight. Regarding opamp bias current (set to ~6.25mA by default). The bias requirements are mostly a function of Q; Fc affects the requirement only by 25% or so. I've listed what I believe are the worst case max input voltage without current limiting.
Maximum safe voltage input
Q=0.5 20 Vrms
Q=0.7 12 Vrms
Q=1.0 7 Vrms
Q=1.4 4.5 Vrms
Q=2.4 2 Vrms
(some of these require rather more supply voltage than +-15V)
If you need more voltage or are concerned for whatever reason, just increase the bias. You can get up to at least 20mA without changing any active components, so I don't think anyone need be concerned about bias.
Maximum safe voltage input
Q=0.5 20 Vrms
Q=0.7 12 Vrms
Q=1.0 7 Vrms
Q=1.4 4.5 Vrms
Q=2.4 2 Vrms
(some of these require rather more supply voltage than +-15V
)If you need more voltage or are concerned for whatever reason, just increase the bias. You can get up to at least 20mA without changing any active components, so I don't think anyone need be concerned about bias.
Hi Tiroth,
thanks! And those plots look really good.
I rethought the grounding and 4.7k resistor thing, I guess my concerns were unfounded.
One more thing: does the MOX board provide for op amp decoupling caps at each IC pin? The Mopsamp may have them onboard but for drop in IC's they would be needed on the MOX board...
thanks! And those plots look really good.
I rethought the grounding and 4.7k resistor thing, I guess my concerns were unfounded.
One more thing: does the MOX board provide for op amp decoupling caps at each IC pin? The Mopsamp may have them onboard but for drop in IC's they would be needed on the MOX board...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.