Wrong question/wrong person, I don't think too much bout caps. Choose foils, any will be very good, if you find any MKP that fit your needs/size, it will be a perfect crossover.
Regards
Flo
MKS or MKT will give surprisingly good performance.
They are small, they are cheap.
They are non polar.
They are available in a wide range of voltages.
They are available in a wide range of values.
They are available in a wide range of packages. This is both good and bad.
Metallised polypropylene film (MKP) is usually considered the next step up.
They are much bigger. They are considerably dearer, but getting cheaper. All the other attributes apply.
Polypropylene film and metal foil is the next step up.
These are very much bigger. These are very expensive.
Using these to experiment in a test crossover would require an enormous budget.
But, once all experimentation is completed, you may have enough space and money to try replacing MKS/MKT/MKP with FKP (film and foil).
Note,
some manufacturers have adopted FKP as part of the product identifier even though they are actually metallised film.
They are small, they are cheap.
They are non polar.
They are available in a wide range of voltages.
They are available in a wide range of values.
They are available in a wide range of packages. This is both good and bad.
Metallised polypropylene film (MKP) is usually considered the next step up.
They are much bigger. They are considerably dearer, but getting cheaper. All the other attributes apply.
Polypropylene film and metal foil is the next step up.
These are very much bigger. These are very expensive.
Using these to experiment in a test crossover would require an enormous budget.
But, once all experimentation is completed, you may have enough space and money to try replacing MKS/MKT/MKP with FKP (film and foil).
Note,
some manufacturers have adopted FKP as part of the product identifier even though they are actually metallised film.
perfect, thanks for the quick answers. I selected vishay CMF55 resistors and vishay mkp caps from farnell. Prices are Ok. Upgrading is always possible. I want to put everything on verboard as close as possible together and close to the xlr outputs.
So i will be building a pre amp with
- The B1 buffers and 2nd order filter as desribed here
- balanced setup
- unbalanced to balanced conversion on some of the inputs
- exernal PSU
- 46 step 4 deck 10K att.
- 21 step 2 deck 20K (stereo) att. for the low and high volume control.
- Maybe an additional sub out using the same crossover and buffers
I am now using unbalanced to balanced converters (from Thel Audioworld) in my otherwise passive pre amp on some of the inputs. These converters use BB OPA-4227. Is there a better (without opamp) for this?
So i will be building a pre amp with
- The B1 buffers and 2nd order filter as desribed here
- balanced setup
- unbalanced to balanced conversion on some of the inputs
- exernal PSU
- 46 step 4 deck 10K att.
- 21 step 2 deck 20K (stereo) att. for the low and high volume control.
- Maybe an additional sub out using the same crossover and buffers
I am now using unbalanced to balanced converters (from Thel Audioworld) in my otherwise passive pre amp on some of the inputs. These converters use BB OPA-4227. Is there a better (without opamp) for this?
Lundahl Transformers LL1570-XL
http://www.4tubes.com/SCHEMATICS/LU...put-transformers/XLR-to-RCA-with-LL1570X0.JPG
http://www.4tubes.com/SCHEMATICS/LU...put-transformers/XLR-to-RCA-with-LL1570X0.JPG
Thanks for the info prooptiki
What about the power supply, in total i have per channel for a symmetric setup:
+ signal B1 - high pass - B1 - high level control - B1
- signal B1 - high pass - B1 - high level control - B1
+ signal B1 - low pass - B1 - high level control - B1
- signal B1 - low pass - B1 - high level control - B1
This yields 12 buffers per channel! I was thinking of one power supply per channel of 12 buffers. Or should i split the PSU also into + and - signal, this giving 4 PSU's, each serving 6 buffers.
I have 10 15000 uF Caps. So i would only need the 3W (1Ohm?) resistors and several PSU's. Any ideas on a good PSU for this? What about 9V batteries?
regards
What about the power supply, in total i have per channel for a symmetric setup:
+ signal B1 - high pass - B1 - high level control - B1
- signal B1 - high pass - B1 - high level control - B1
+ signal B1 - low pass - B1 - high level control - B1
- signal B1 - low pass - B1 - high level control - B1
This yields 12 buffers per channel! I was thinking of one power supply per channel of 12 buffers. Or should i split the PSU also into + and - signal, this giving 4 PSU's, each serving 6 buffers.
I have 10 15000 uF Caps. So i would only need the 3W (1Ohm?) resistors and several PSU's. Any ideas on a good PSU for this? What about 9V batteries?
regards
Thanks for the info prooptiki
What about the power supply, in total i have per channel for a symmetric setup:
+ signal B1 - high pass - B1 - high level control - B1
- signal B1 - high pass - B1 - high level control - B1
+ signal B1 - low pass - B1 - high level control - B1
- signal B1 - low pass - B1 - high level control - B1
This yields 12 buffers per channel! I was thinking of one power supply per channel of 12 buffers. Or should i split the PSU also into + and - signal, this giving 4 PSU's, each serving 6 buffers.
I have 10 15000 uF Caps. So i would only need the 3W (1Ohm?) resistors and several PSU's. Any ideas on a good PSU for this? What about 9V batteries?
regards
Hi,
with some savety margin, calculate 20mA per B1 and add some extra margin to the sum, not too much. Each B1 draws a constant current, so there is no "need for speed", you just need some stability.
You can split the supplies the way you prefer. You can use one, two four or 24 for your setup. If you use more than one, just look for proper grounding.
No it does not.
The B1 is a source follower loaded with a constant current sink.
If the quiescent current is say 10mA then the maximum output is ~+-9mApk.
The supply current will vary from ~1mA to 19mA if the FET varies the full +-9mA. If the output current is +-1mA then the supply sends out 10mA+-1mA, i.e. 9mA to 11mA. It is NOT constant.
Where do you find your information?
Do you ever query it?
Sorry, my post was written badly, shoud stop posting when too tired...
According to my experience the sound of a B1 gets worse if the load gets below 5k for line-level voltages. I think, I wrote this some times earlier in this thread, it corresponds to the output impedance of 0.5k-1k, depending on the specific design. So the maximum current that should be drawn from the B1 is about 0.5mA (peak).
As a consequence current varies from 9.5mA to 10.5mA which - in my opinion - can be regarded as (nearly) constant especially when designing a power supply with a sufficient savety margin.
Sorry that my post was too unclear.
I am thinking if this "type" of B1 would give the same results, so as to avoid caps...
Except for the values of the Rs that are not that critical, for a single B1 you should not go below 220R for R4 to get a stable Buffer, it's the same as the ones posted above. The symmetric-supply version works without caps, but in my opinion it's save to leave them in place because even if your XO does not have any dc-offset, you might use a source that generates some offset in the future and a power amp without any DC-blocking.
Best regards
Flo
Thanks, I am already using the classic B1 buffer + F5 when I start the b1 I get a huge offset about 1.5v (the f5 is on) after 35sec the offset goes to zero (is this logical?)! So I have to make a relay delay circuit... Or there is something better? I am anxious all the time if the electical power go off....!
Someone wrote in the b1 buffer thread:
"-Change output cap from 10uf MKP type to electrolytic, offset increases to 900mV from cold and decreases very slow
-Change output cap from 10uf to 4.7uf SCR MKP, offset starts around 30mV and declines quite faster than 10uf MKP cap."
My B1+jfet matched came from passdiy, 50k pot, for output cap I use 3 parallel caps (3.3+3.3+4.7 = 11.3uf the cheap ones from Mundorf and Audyn mkp)
I read also somewhere that we can change the 10uf cap to 4.7uf plus one resitor with the same very good result but I can not find it....
Last edited:
"Originally posted by Babowana
With 10uF+1K into 10k load, I estimate the low frequency roll-off at 1.5Hz.
With 10uF+1K into 47k load, I estimate the low frequency roll-off at 0.4Hz.
In the similar way as flws:
With 2.2uF+4.7K into 10k load, I estimate the low frequency roll-off at 5Hz.
With 2.2uF+4.7K into 47k load, I estimate the low frequency roll-off at 1.5Hz.
My conclusion is that you could use the combi of 2.2uF+4.7K if you do not mind the 5Hz roll-off. "
Is this correct?
With 10uF+1K into 10k load, I estimate the low frequency roll-off at 1.5Hz.
With 10uF+1K into 47k load, I estimate the low frequency roll-off at 0.4Hz.
In the similar way as flws:
With 2.2uF+4.7K into 10k load, I estimate the low frequency roll-off at 5Hz.
With 2.2uF+4.7K into 47k load, I estimate the low frequency roll-off at 1.5Hz.
My conclusion is that you could use the combi of 2.2uF+4.7K if you do not mind the 5Hz roll-off. "
Is this correct?
Hi,
I don't know the LDR in practice. In my opinion, you can use whatever you want. The Pot (or whatever) after the B1 should be an impedance of at least 5k. 20k would be better.
I find it practical to have defined steps (e.g. 1dB) for the high and low attentuation. Your settings will be easlily reproducable then and it will be comfortable to change e.g. amps if you know their gain.
Best regards
I don't know the LDR in practice. In my opinion, you can use whatever you want. The Pot (or whatever) after the B1 should be an impedance of at least 5k. 20k would be better.
I find it practical to have defined steps (e.g. 1dB) for the high and low attentuation. Your settings will be easlily reproducable then and it will be comfortable to change e.g. amps if you know their gain.
Best regards
Harold asked about the hi- and low- Attentuation after the hi- resp. lowpass of the filter of the crossover we are talking about in this thread. I think it would be wise to put a buffer in between.
A buffer goes AFTER a source that cannot adequately drive the cables and receiver.
If you add an attenuator and that attenuator cannot drive the subsequent output then the buffer is after the attenuator not before and not between it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.