Here's the method I use for running my DATSv3 natively on my Mac using REW.
Member
Joined 2003
Handy instruction, really this can apply to any OS where someone would want to use REW with the DATS.
The use of 10k sense resistor is a surprising choice from Dayton, maybe the circuit has more going on than a simple voltage divider, but 10k would make for most of the signal dissipated by the sense resistor, very little signal across the woofer, no wonder people often get noisy measurements with the WT/DATS.
The use of 10k sense resistor is a surprising choice from Dayton, maybe the circuit has more going on than a simple voltage divider, but 10k would make for most of the signal dissipated by the sense resistor, very little signal across the woofer, no wonder people often get noisy measurements with the WT/DATS.
10kΩ is just about right, and it seem like a pretty clever design: the DATS is assuming the output is a (constant) current source which is holds as long as the load impedance is no higher than 1/10th the output impedance. This gives the measured device impedance an upper limit of about 1kΩ, which is can cover drivers with peaks that extend into the hundreds of ohms. From the PCM2900 data sheet it looks like the ADC is 16-bit with a SNR of 89dB meaning there's at least 14-bits of precision to play with, which is plenty to get 1% accuracy with precision down to tens of milliohms. By definition, to determine TSPs (or small signal parameters) you want to measure the impedance with a small signal across the speaker terminals otherwise the linear assumptions/equations don't hold up.
If you're having issues with of noisy signals, make sure to measure in a quiet environment (when using REW you can see how the DATS picks up enviromental/ambient noise). Another thing that helps is using a longer sweep along with averaging multiple sweeps to reduce random noise. I'm able to get consistent and accurate using the settings shown as long as it's not too loud. I don't really remember using the DATS SW, and can't remember if there's an option to automatically average multiple sweeps.
If you're having issues with of noisy signals, make sure to measure in a quiet environment (when using REW you can see how the DATS picks up enviromental/ambient noise). Another thing that helps is using a longer sweep along with averaging multiple sweeps to reduce random noise. I'm able to get consistent and accurate using the settings shown as long as it's not too loud. I don't really remember using the DATS SW, and can't remember if there's an option to automatically average multiple sweeps.
Member
Joined 2003
I don't use DATS, I used to have a WT3 and while it was "fine", the major problem was poor/weak alligator clip connections. I use my own resistor jig now with REW or ARTA, but sense resistor is only 10 ohm. I should give it a try with a high value resistor and see how much more noisy the signal becomes. Could be a simple solution to the "constant voltage/current" source problem, where 10 ohm sense resistor provides quite a variable voltage to the driver over frequency, however I have found the results don't affect cabinet model so it's more of a "nice to have" feature.
My comment on noise with WT/DATS is just from common problems posted by many users, it seems every week or two there is someone somehwere having trouble with it, but often related to volume control settings in windows. Even with your own meaurements, as you mention there is some noise especially at low frequency that requires some averaging to get a more reliable result.
My comment on noise with WT/DATS is just from common problems posted by many users, it seems every week or two there is someone somehwere having trouble with it, but often related to volume control settings in windows. Even with your own meaurements, as you mention there is some noise especially at low frequency that requires some averaging to get a more reliable result.
The idea is to use a low value sense resistor like 0.1Ω if you are approximating a voltage source (assuming constant voltage), or a high value like 10kΩ if you are assuming a constant current. If you do something in between (like 10Ω) you cannot make assumptions of constant current nor constant voltage, and it introduces electrical damping to the system. 10Ω is perfect if you are running a single ended tube amp that has a 10Ω output impedance 🙂
In theory it's better to measure with a low value sense resistor which better approximates the vast majority of amplifiers which act as voltage sources, but in practice I just use a DATS which is close enough for most of the stuff I'm doing.
In theory it's better to measure with a low value sense resistor which better approximates the vast majority of amplifiers which act as voltage sources, but in practice I just use a DATS which is close enough for most of the stuff I'm doing.