A 2.5 way basreflex loudspeaker floorstander with crossover was tested.
Multiple measurements are consistent and only differ at first impedance peak. 1 Ohm difference. ..Allways... what does this mean? Resister measurements outcomes between the 2 measure methods are the same.
Resister measurements that were done showed expected values and are within range for both methods.
Comparison measurement DATSv3 versus DATSv3 + REW
DATSv3 + REW Red
DATSv3 Purple
In REW all values except “the leads resitance value” for the impedance setup were take from the manual DATv3 on a Mac written by olen.
10kΩ for the Rsense value. (based on the value of output resistor) Don’t know my Dats R value
Set Rsense = 10kΩ
100kΩ for Rin (where is this value derived from?)
10mΩ for Rleads. I used 0.7399 Ohm Messured earlier with dats sofware.
For the reference cal I used a 4.952 ohm resistor..
-- INFO -- My version of the DATSv3 is different from olen’s...See earlier post with picture...If my dats series output resistor value differs say 2000 Ohm from the 10000 in olens Manual could that alter the reponse, and then ONLY doing so in the left part of the graph we are seeing?
Thanks I appreciate your input and thoughts..
OS on laptop, Manjaro Linux with Wine installed....Linux version of REW 5.31.3 installed
install
DATS_V3_setup.exe installed with wine installer. (right click install with Wine windows program Loader)
Installation path: home/.wine/drive_c/Program Files/
Open console and type winecfg hit ENTER
In Wine configuration window that pops up click Add application. Go to c/Program Files/DATSV3/ and select en add DATS.exe
In Wine configuration window select the just added DATS.exe and click audio at top window
I chose windows version 7 but other OS may work.
For the input and output device now there’s the just installed PCM2900C Audio CODEC Analog stereo option to choose. The rest i left at “System default” cause other settings also didn’t add any specific in and outputs. In other words I CAN NOT CHOOSE “SPEAKER” OR “LINE IN” like olen could in the manual DATSv3 on a MAC.
Also in “PulseAudio Volume Control” I chose PCM2900C Audio CODEC profile: analog stereo Duplex
When in REW ...for input and output device same CODEC is chosen...in and output were left to default
No option to CHOOSE “SPEAKER” OR “LINE IN” I think I have might have to edit some settings in Manjaro or ad pipewire or jack to route audio ins and outs...If you guys think the measurement is ok I leave it as it is for now.
have a good weekend
Multiple measurements are consistent and only differ at first impedance peak. 1 Ohm difference. ..Allways... what does this mean? Resister measurements outcomes between the 2 measure methods are the same.
Resister measurements that were done showed expected values and are within range for both methods.
Comparison measurement DATSv3 versus DATSv3 + REW
DATSv3 + REW Red
DATSv3 Purple
In REW all values except “the leads resitance value” for the impedance setup were take from the manual DATv3 on a Mac written by olen.
10kΩ for the Rsense value. (based on the value of output resistor) Don’t know my Dats R value
Set Rsense = 10kΩ
100kΩ for Rin (where is this value derived from?)
10mΩ for Rleads. I used 0.7399 Ohm Messured earlier with dats sofware.
For the reference cal I used a 4.952 ohm resistor..
-- INFO -- My version of the DATSv3 is different from olen’s...See earlier post with picture...If my dats series output resistor value differs say 2000 Ohm from the 10000 in olens Manual could that alter the reponse, and then ONLY doing so in the left part of the graph we are seeing?
Thanks I appreciate your input and thoughts..
OS on laptop, Manjaro Linux with Wine installed....Linux version of REW 5.31.3 installed
install
DATS_V3_setup.exe installed with wine installer. (right click install with Wine windows program Loader)
Installation path: home/.wine/drive_c/Program Files/
Open console and type winecfg hit ENTER
In Wine configuration window that pops up click Add application. Go to c/Program Files/DATSV3/ and select en add DATS.exe
In Wine configuration window select the just added DATS.exe and click audio at top window
I chose windows version 7 but other OS may work.
For the input and output device now there’s the just installed PCM2900C Audio CODEC Analog stereo option to choose. The rest i left at “System default” cause other settings also didn’t add any specific in and outputs. In other words I CAN NOT CHOOSE “SPEAKER” OR “LINE IN” like olen could in the manual DATSv3 on a MAC.
Also in “PulseAudio Volume Control” I chose PCM2900C Audio CODEC profile: analog stereo Duplex
When in REW ...for input and output device same CODEC is chosen...in and output were left to default
No option to CHOOSE “SPEAKER” OR “LINE IN” I think I have might have to edit some settings in Manjaro or ad pipewire or jack to route audio ins and outs...If you guys think the measurement is ok I leave it as it is for now.
have a good weekend
To verify that DATS worked correctly in linux Manjaro (without using REW) I ran same test in windows 7 where i could set speaker and line-in inputs and outputs for the AUDIO CODEC devices. Imported the DATS 'manjaro" and "W7" freshly made impedance zma files in REW (witch also is installed on windows 7) and they match! So thats ok...Then ran REW + DATS also in w7 and that shows same result as in Manjaro. The same Larger impedance peak at arround 28hz for REW. 1.2 Ohm higher than DATS....I'll try to recalibrate inpedance in REW with other values for Rsense and R-in and see what it does to the measurements. Since i have a different DATSv3 then the @olen manual shows it might be worth a try.
Hi John, sorry about the delayed response, it was a busy holiday weekend.
I bought another DATSv3 earlier this year but didn’t take it apart. I didn’t need to because newer versions of REW have built in calibration steps that work perfectly. I use a lower value resistor (3 to 10 ohms) as the known resistor value, which is closer to the impedance of the speakers I’m testing. I use a calibrated 4-wire DMM to measure the reference resistor with milliohm resolution. I’d be happy to send you a calibrated resistor or two if you send me you me your address in a DM.
The impedance plot you show is characteristic of a ported speaker (or passive radiator), the magnitude of the first impedance peak can change with drive level or differences in air load on the port. Did you move the speaker between tests? Is the port partially blocked or near a wall? Were the drive levels consistent between test?
As an aside, one way to test port compression is to measure electrical impedance at increasing drive levels, you’ll see the first impedance peak start to flatten out. It might be difficult to see this with a DATS because it can’t provide enough output power to drive a speaker port into compression.
Hopefully this gives you some ideas to answer your question.
I bought another DATSv3 earlier this year but didn’t take it apart. I didn’t need to because newer versions of REW have built in calibration steps that work perfectly. I use a lower value resistor (3 to 10 ohms) as the known resistor value, which is closer to the impedance of the speakers I’m testing. I use a calibrated 4-wire DMM to measure the reference resistor with milliohm resolution. I’d be happy to send you a calibrated resistor or two if you send me you me your address in a DM.
The impedance plot you show is characteristic of a ported speaker (or passive radiator), the magnitude of the first impedance peak can change with drive level or differences in air load on the port. Did you move the speaker between tests? Is the port partially blocked or near a wall? Were the drive levels consistent between test?
As an aside, one way to test port compression is to measure electrical impedance at increasing drive levels, you’ll see the first impedance peak start to flatten out. It might be difficult to see this with a DATS because it can’t provide enough output power to drive a speaker port into compression.
Hopefully this gives you some ideas to answer your question.
Thanks @olen.No problem, and thats a very kind offer you did! thank you.Luckily i got a electronics shop nearby. For Cal in REW i used an approx 0.6 watt (guess) metal-film 1% resister 4.955 ohms. Earlier i did the cal with 0.5W 0.1% 100 ohms. Results were almost identical.
a) did not move speakers
b) no blocked port (unchanged constant distances to objects...at least 1m) bb) did tests with my hand en cloth to block the port slightly and the graph changes big time.If hand is close to port the graph changes allready.🙂
c) constant temperature 17.5 C (the measurements taken closest together in time were used)
d) i can improve on matching drive levels..and sweep length...although deviations in the graph were small, with the different levels used .They didnt result in a 1.2 ohms diff in measurement using one method of measuring..
Good to know! "with higher levels first impedance peak could flatten out". Levels have to be considerably higher then the 10db DATS max for the bigger 150W nominal speaker to show significant compression i presume.But ok compression levels maybe gradual rising and starting early at medium levels i don't know.Im not experienced. I will take new measurements for the BIG speaker later with spl meter.
I did this spl comparison reading for a small cheap desktop bassreflex. -4.9db in REW was equal to DATS level 1 sweep....(not shure if one can set the level in DATS when using a single sweep (not using the function "linearity test") Even at unmatching levels the graph looked roughly alike at peaks.Thats because these cheap very low quality ones not go as low as the big speakler i presume.The bigger Heco speakers are also harder to drive.
olen,
Is a 1W non-inductive ref resister sufficient for Cal use in REW+DATS? or will a 0.6W be as good?
1) Do you use a beta version with even better built in cal steps?. I'm using REW 5.31.3. not shure if i can use the beta in linux tho, due to the not built-in Java.
2) are you still using the values Rsense 10k (thats the max in my REW) and Rin 100k ?
b Would it be usefull to measure the output impedance of the DATS and use that for Rsense or would that nitpicking?
3) are you using very low ohms 'leads" that didnt come with the DATS olen.My leads are 0.737ohms measured with DATS. You entered 0.010 ohms
4)in relation to compression (port) and a impedance graph seen at different levels. If i understand you correctly... a flattening out of the first peak at high enough levels could mean compression.
5) do you use a impedance rig and amp with rew to test at higher levels?
Here's the "cheap Speakers pics".Second pic zoomed in... DATS solo in red vs DATS+REW. Same level. The deviation second peak/top was constant...maybe due to noise...averaging doesnt really improve on peaktop, REW 1/48 smoothing. Perhaps REW sofware has higher resolution ha.
probably 128k / 256 length in rew and 64k in DATS....aaaarrggg Dont know yet if one can set sweep length in DATS.
Ok a lot of questions i hope you don't mind. No hurry Respond as you wish. Bye have a good week!
a) did not move speakers
b) no blocked port (unchanged constant distances to objects...at least 1m) bb) did tests with my hand en cloth to block the port slightly and the graph changes big time.If hand is close to port the graph changes allready.🙂
c) constant temperature 17.5 C (the measurements taken closest together in time were used)
d) i can improve on matching drive levels..and sweep length...although deviations in the graph were small, with the different levels used .They didnt result in a 1.2 ohms diff in measurement using one method of measuring..
Good to know! "with higher levels first impedance peak could flatten out". Levels have to be considerably higher then the 10db DATS max for the bigger 150W nominal speaker to show significant compression i presume.But ok compression levels maybe gradual rising and starting early at medium levels i don't know.Im not experienced. I will take new measurements for the BIG speaker later with spl meter.
I did this spl comparison reading for a small cheap desktop bassreflex. -4.9db in REW was equal to DATS level 1 sweep....(not shure if one can set the level in DATS when using a single sweep (not using the function "linearity test") Even at unmatching levels the graph looked roughly alike at peaks.Thats because these cheap very low quality ones not go as low as the big speakler i presume.The bigger Heco speakers are also harder to drive.
olen,
Is a 1W non-inductive ref resister sufficient for Cal use in REW+DATS? or will a 0.6W be as good?
1) Do you use a beta version with even better built in cal steps?. I'm using REW 5.31.3. not shure if i can use the beta in linux tho, due to the not built-in Java.
2) are you still using the values Rsense 10k (thats the max in my REW) and Rin 100k ?
b Would it be usefull to measure the output impedance of the DATS and use that for Rsense or would that nitpicking?
3) are you using very low ohms 'leads" that didnt come with the DATS olen.My leads are 0.737ohms measured with DATS. You entered 0.010 ohms
4)in relation to compression (port) and a impedance graph seen at different levels. If i understand you correctly... a flattening out of the first peak at high enough levels could mean compression.
5) do you use a impedance rig and amp with rew to test at higher levels?
Here's the "cheap Speakers pics".Second pic zoomed in... DATS solo in red vs DATS+REW. Same level. The deviation second peak/top was constant...maybe due to noise...averaging doesnt really improve on peaktop, REW 1/48 smoothing. Perhaps REW sofware has higher resolution ha.
probably 128k / 256 length in rew and 64k in DATS....aaaarrggg Dont know yet if one can set sweep length in DATS.
Ok a lot of questions i hope you don't mind. No hurry Respond as you wish. Bye have a good week!
Hi @John64
1. Power rating should not make much of a difference at the levels DATS tests, to see if there’s any effect: after you complete calibration run multiple impedance sweeps and see if there is any change (thermal drift), technically a larger, higher rated should have more margin for thermal stability. The tests levels are so low that 0.6W or even lower should have plenty of thermal margin. DATS is intended to measure impedance for determining small signal parameters, so….
2.No, I use whatever value the calibration step determines, I only used the 10k back before the calibration step existed.
3. I’m using the stock leads, 0.737ohms does seem a bit high, as the output is a current source with series with a 10k resistor I guess an extra ohm doesn’t make a difference. If you have an extra set of test leads I would try them and see, you’ll need to recalibrate. I would expect less than 100 milliohms for a pair of shorted test leads.
4. Yes that the correct understanding, I’ll see if I can pull up an example
5. I use an AP or a klippel to run high power impedance sweeps. I don’t personally own a klippel system, but it is the easier way. I have an AP 2722 and a very large 100-milliohm /250W resistor that I use for a current shunt. A large, low output impedance amplifier drives the shunt resistor in series with the speaker(DUT). I run a sweep measuring 2 channels simultaneously: the voltage across the DUT and the voltage across the shunt. I amplify/multiply the voltage measured by the shunt by 10x (or 20dB) to get the units into Amperes, and divide the voltage channel by the current channel to get a result in ohms. Another possibly better way is to use a current probe for the current measurement, assuming the noise floor and range are sufficient, but a good current probe tends to be expensive.
Thanks @olen. 🙂
2) In your manual 'dats on a mac' with rew, you used 10kΩ for the Rsense value based on the value of the output series resistor of the DATS.This was part of the calibration proces you did in the manual. Not shure when you now say 'i only used 10k back before the cal proces existed'
b) still have to enter a Rsense value for REW 5.31.3 during 3 step calibration.
Hope you saw the pic with detailed print-layout with components of my Dats..mine's is different.See below. Sinse i dont know the dats output series resister is there a way to find out (measure it) or doesnt it matter? .Perhaps all DATS output series resistance have a high value near 10k?
2) In your manual 'dats on a mac' with rew, you used 10kΩ for the Rsense value based on the value of the output series resistor of the DATS.This was part of the calibration proces you did in the manual. Not shure when you now say 'i only used 10k back before the cal proces existed'
b) still have to enter a Rsense value for REW 5.31.3 during 3 step calibration.
Hope you saw the pic with detailed print-layout with components of my Dats..mine's is different.See below. Sinse i dont know the dats output series resister is there a way to find out (measure it) or doesnt it matter? .Perhaps all DATS output series resistance have a high value near 10k?
Hmm, sorry for the confusion, I found on more recent versions of REW if you use the built-in calibration procedure you don’t need to manually enter these values, I don’t remember exactly when the cal procedure changed. When I initially started using REW with my DATS the self cal procedure was not in place, it wasn’t until much later that I documented the steps. I continued using the manual method that had been working and didn’t try the self cal steps until later. I recommend using the self calibration, maybe try it both ways and compare the measurement using your reference resistors - my guess is using the built in cal procedure provides more accurate measurement results.
Hi @olen. Thanks! .(almost confusion again..for a moment i thought you now talk about the simple self calibration of a dats) ..Imp Calibration for REW..well thats strange, because the red circlelled part in the manual about cal procedure is still the same in my 5.31.3v of REW. And this is the latest non-beta version!
1) With your "rew built in self calibration" procedure .Did the rew screen look the same as circleled in red with the difference that the 10k, 100k values were pre-filled in? is that what you mean by "rew built in self calibration"?
2) With "rew built in self calibration" you didn't have to enter a leads value? i assume it still mandatory to enter a leads value.
3)For the reference cal i am assuming its still prefered to enter a resister value that is close to what is to be expected to measure?
For example: value 4 ohm when a 5 ohm speaker is measured.
i can remember now the 10k was already pre filled in. Maybe the 100k was also filled in.I dont remember...
We are probably talking in circles and i've been using the built-in cal procedure all along?. I was thinking the "new" 'built-in cal precedure" you talk about would look much different with no 3 steps to go through.confusion... wish english was my native language
1) With your "rew built in self calibration" procedure .Did the rew screen look the same as circleled in red with the difference that the 10k, 100k values were pre-filled in? is that what you mean by "rew built in self calibration"?
2) With "rew built in self calibration" you didn't have to enter a leads value? i assume it still mandatory to enter a leads value.
3)For the reference cal i am assuming its still prefered to enter a resister value that is close to what is to be expected to measure?
For example: value 4 ohm when a 5 ohm speaker is measured.
i can remember now the 10k was already pre filled in. Maybe the 100k was also filled in.I dont remember...
We are probably talking in circles and i've been using the built-in cal procedure all along?. I was thinking the "new" 'built-in cal precedure" you talk about would look much different with no 3 steps to go through.confusion... wish english was my native language
Your English is infinitely better than my Dutch 🙂 Apologies for my vague descriptions
1) Unfortunately I don’t remember and didn’t note what it looked like previously. When I use the REW’s built in cal procedure I don’t enter any values other than the exact measured value of my reference resistor. For example 4.321 ohms. After the calibration process is complete I believe REW fills in these other values.
2) I don’t believe you need to enter any value for the leads, I believe this value is taken care of when you run the cal test with the leads shorted.
3) Yes that is correct. You want to use a reference resistor that is in the neighborhood of the values you are trying to measure. I use a reference resistor where it’s specific value was measured on a calibrated micro-ohm meter - in an attempt to achieve 4+ digits of precision.
1) Unfortunately I don’t remember and didn’t note what it looked like previously. When I use the REW’s built in cal procedure I don’t enter any values other than the exact measured value of my reference resistor. For example 4.321 ohms. After the calibration process is complete I believe REW fills in these other values.
2) I don’t believe you need to enter any value for the leads, I believe this value is taken care of when you run the cal test with the leads shorted.
3) Yes that is correct. You want to use a reference resistor that is in the neighborhood of the values you are trying to measure. I use a reference resistor where it’s specific value was measured on a calibrated micro-ohm meter - in an attempt to achieve 4+ digits of precision.
@olen thanks! English hmmm i try my best to be clear but it takes some effort. That's ok still learning but my brain is not getting younger.
Ok i will reset the imp call in REW to see what the values look like in rew's fields...big chance the 10k and 100k after cal-clear wil revert back to what's been used lately.Namely also 10k and 100k
1) Do you use REW beta version?
2) leads value....I believe this value is taken care of when......makes sense yes (if i got the right rew version) Resisters measured in REW after earlier cal procedure WITH 0.74ohm entered for leads gave me correct values for the then measured resisters .....i'll test it....lost my tiny 4.9ohm resister...buy some new ones later.have a good weekend.Cheers
Ok i will reset the imp call in REW to see what the values look like in rew's fields...big chance the 10k and 100k after cal-clear wil revert back to what's been used lately.Namely also 10k and 100k
1) Do you use REW beta version?
2) leads value....I believe this value is taken care of when......makes sense yes (if i got the right rew version) Resisters measured in REW after earlier cal procedure WITH 0.74ohm entered for leads gave me correct values for the then measured resisters .....i'll test it....lost my tiny 4.9ohm resister...buy some new ones later.have a good weekend.Cheers