• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

New Denon 3700/4700 DAC chip

OP
M

mike7877

Addicted to Fun and Learning
Joined
Aug 5, 2021
Messages
698
Likes
140
So you couldn't even level match?

I have a scope so could match levels, but it's overkill when the other steps are done by ear. Comparing noise, the difference between my DAC and the 3700's is notable. Playing silence through the 3700's DAC with the receiver's volume at 0db, noting the noise, then switching the input to the external DAC and increasing the volume control until it's hiss is at the same level is a method for estimating (in direct/pure direct).

My DAC: -102db,
3700 v1: -98db.
Difference: 4db.

Where should I have had to turn the volume of my DAC to for the levels of hiss to match?
+4db

Where did I have to turn it?
+16

Subtract it from my DAC's known THD+n (which is basically just n, my DAC's harmonics are lower in level than n)

102 - 16 = 86

86 isn't 98.

Is it exact? No, but it's probably close. It's also a best case, it assumes the 3700 also has all harmonics lower inlevel than noise like my DAC.

Now, using the same method (volume control), the noise in Auto is 12db higher than Direct mode.

86-12 =74

86 isn't horrible, but 74 is.

Why aren't there ANY measurements here with the processing turned on? Most people use their home theater equipment with room correction! There's no, I repeat: NO reason for ANY of the noise added with processing, 12db in the 3700s case. It's all digital! Are the engineers retarded? Yes, I think they must be.

I bought this 3700 thinking 98db snr!

Turns out, it's probably 74.

Should I be subtracting 10-15db from all the reviews on this site?

I can't believe what garbage everything is. Twenty flipping five db more noise than anticipated. In other words: a few hundred times more noise than anticipated. I'm livid
 
Last edited:

Chrispy

Master Contributor
Forum Donor
Joined
Feb 7, 2020
Messages
7,945
Likes
6,105
Location
PNW
I have a scope so could match levels, but it's overkill when the other steps are done by ear. Comparing noise, the difference between my DAC and the 3700's is notable. Playing silence through the 3700's DAC with the receiver's volume at 0db, noting the noise, then switching the input to the external DAC and increasing the volume control until it's hiss is at the same level is a method for estimating (in direct/pure direct).

My DAC: -102db,
3700 v1: -98db.
Difference: 4db.

Where should I have had to turn the volume of my DAC to for the levels of hiss to match?
+4db

Where did I have to turn it?
+16

Subtract it from my DAC's known THD+n:

102 - 16 = 86

86 isn't 98.

Is it exact? No, but it's probably close.
Using the same method, the noise in Auto is 12db higher than Direct.

86-12 =76

86 isn't horrible, but 76 is.

Why aren't there ANY measurements here with the processing on? Most people use it in home theater. There's no reason for any of the 12db of added noise, the processing is all digital!

I bought this 3700 thinking 98db snr!

Turns out, it's probably 76

Should I be subtracting 10-15db from all the reviews on this site?

I can't believe what garbage everything is
I think you're likely overthinking it myself or focusing on things that really don't particularly matter or relate IMHO. So use the scope and report back....
 
OP
M

mike7877

Addicted to Fun and Learning
Joined
Aug 5, 2021
Messages
698
Likes
140
I think you're likely overthinking it myself or focusing on things that really don't particularly matter or relate IMHO. So use the scope and report back....

What doesn't matter or relate? I haven't overthought anything, I put things succinctly.
 

Chrispy

Master Contributor
Forum Donor
Joined
Feb 7, 2020
Messages
7,945
Likes
6,105
Location
PNW
What doesn't matter or relate? I haven't overthought anything, I put things succinctly.

Well relating your experience to the spec/measurments without actually measuring for one. Using different inputs/processing in an avr could have different results for another unless you at least level match for audibility issues. The particular chip used another. YMMV but seems mostly just your own subjective experience/opinion so far.
 
OP
M

mike7877

Addicted to Fun and Learning
Joined
Aug 5, 2021
Messages
698
Likes
140
Well relating your experience to the spec/measurments without actually measuring for one. Using different inputs/processing in an avr could have different results for another unless you at least level match for audibility issues. The particular chip used another. YMMV but seems mostly just your own subjective experience/opinion so far.

I can level match by ear to 1db, I have experience designing loudspeaker crossovers. I also know how to configure this receiver properly

"The particular chip used another"
You don't make sense.

Maybe 3700 v2 has 86db snr dac, maybe 83, maybe 89. I'm not doing a tonne of work when I already know it's about 86, maybe worse.

Then if you want processing it's lowered further - about 12db!

Why are there no measurements on this site with commonly used sound processing? Would all receivers look like the complete garbage they are, untouchable, not worth half of what they're charging?
 

Chrispy

Master Contributor
Forum Donor
Joined
Feb 7, 2020
Messages
7,945
Likes
6,105
Location
PNW
I can level match by ear to 1db, I have experience designing loudspeaker crossovers. I also know how to configure this receiver properly

"The particular chip used another"
You don't make sense.

Maybe 3700 v2 has 86db snr dac, maybe 83, maybe 89. I'm not doing a tonne of work when I already know it's about 86, maybe worse.

Then if you want processing it's lowered further - about 12db!

Why are there no measurements on this site with commonly used sound processing? Would all receivers look like the complete garbage they are, untouchable, not worth half of what they're charging?

1 dB is insufficient. Comparing different features in an avr aren't all that meaningful generally IMO. AVRs aren't "horrible" compared to my old 2ch separates at least in terms of audible performance (but old gear has no dac at all of its own particularly outside of sources). It would be nice to see more actual use measurements of all aspects of an avr, tho. Your post started with what chip is being used which is somewhat irrelevant without consideration of implementation. I still think you're overthinking it or just too damn picky :)
 
OP
M

mike7877

Addicted to Fun and Learning
Joined
Aug 5, 2021
Messages
698
Likes
140
The 3700's DAC has approximately 16db more noise than the external DAC I plugged into the analog input.

If the external DAC is given 18db more gain than the 3700's internal DAC, the external dac's noise sounds obviously louder - very obviously louder. If it's given only 14db more gain, it's obviously quieter - unmistakably quieter.

So the DAC in the 3700 has between 15 and 17db, probably 16db more noise than the DAC I attached.

The DAC I attached has a THD+n of -102db

102-16 = 86

Tell me what you don't understand or think I'm doing wrong and I'll clear it up.
 

ShadowFiend

Member
Joined
Dec 20, 2020
Messages
69
Likes
89
I said that more than once before and will say this again, that such claims are not credible. If true, that's bad news in terms of measurements.
I think it seems credible. No way any manufacture will change from a cheap 8-channel-DAC Ak4458 to something like ES9018, which is more expensive and require new R&D in such short time for low price products. And there is also a shortage of ESS chips right now.

PCM5012 is decently good, cheap & plenty as Texas Instrument has much better production capability than ESS. Denon also has experiences with PCM5102 for a long time. So as a big manufacture like Denon, PCM5102 is much safer & sensible choice, especially for bread-and-butter products ( 4700 or 6700 AVR). For the more expensive products, they can use better DACs like ES9018 & their remaining AK4490.
 

peng

Master Contributor
Forum Donor
Joined
May 12, 2019
Messages
5,757
Likes
5,328
I think it seems credible. No way any manufacture will change from a cheap 8-channel-DAC Ak4458 to something like ES9018, which is more expensive and require new R&D in such short time for low price products. And there is also a shortage of ESS chips right now.

PCM5012 is decently good, cheap & plenty as Texas Instrument has much better production capability than ESS. Denon also has experiences with PCM5102 for a long time. So as a big manufacture like Denon, PCM5102 is much safer & sensible choice, especially for bread-and-butter products ( 4700 or 6700 AVR). For the more expensive products, they can use better DACs like ES9018 & their remaining AK4490.

First of all, I don't remember the ES9018 being mentioned by anyone and, understandably so as the cost for those would be prohibitive for use in any AVP/C/AVRs. So it is easy to agree with you that there is no way any manufacturer would do such a switch from a mid range IC to a top reference class one.:D

Other than that, I understand your point about the 5102 as a possibility though for reasons I cited I still think it would be very unlikely. Also, if it is credible, then how come there has been no posted links to the source of information directly from Denon?

On the other hand, we do have info from Denon on their Japan website and that's why I don't think the claim of the PCM5102 is credible, again, unless someone post links to info from Denon or Marantz.

I hope you don't mind reading the following as I would like to know if you see what I see, that while the board definitely looks different than the one for the AK4458s but there appears to be still two 48 pin DAC ICs on board; and there's no sign of 7 to 8 PCM5102 (for the main zone) plus two more PCM5100/or 5101 (for the secondary zones):

(8) Denon Replaces AKM AK4458 DAC IC in X4700H and X6700H | Page 6 | Audio Science Review (ASR) Forum

If you magnify the photo of the DAC board you can see the 2 DAC ICs that seem to have the same pin layout, or as least the same pin numbers as the AK4458. It is not possible to tell the model/part numbers, other than we already know it is a replacement of the AK4458. I assume there are other DAC ICs that also have 48 pins, at least the ES9006 would be one of them.

So based on pin numbers that can be seen on the DAC board posted on Denon.jp, I do not think it is likely to be the PCM5102. I hope it would be one that AKM is still producing or have a lot of inventory, and is comparable to the AK4458 in specs. If not, then I guess the next best would be the ES9006 that is only 4 dB lower in SINAD than the AK4458.

As an aside, the following video (thanks to bigguyca) on the SR8015 that has the AK4490 replacement DAC:
Denon Replaces AKM AK4458 DAC IC in X4700H and X6700H | Page 7 | Audio Science Review (ASR) Forum

If you fast forward to the 26 minute point you will see that Marantz has used the ES9010K2M. That is downgrade, in terms of the SINAD spec (6 dB lower), from the AK4490. So since D+M has already used the ESS DAC ICs in their higher models, it is not inconceivable that they may use the ES9006 on the lower models such as the AV7706, SR7015 and SR6015.

Regarding the ES9018, that's a reference class DAC IC that I highly doubt you would find any of them in AVP/C and AVRs regardless of price. The highest ESS DAC ICs used in flagship class AVP/C/AVRs seemed to be the ES9026Pro that is more comparable to the AK4490 but not quite at the level of the ES9018/9028/9038 or even the old 9008.

Unfortunately we are all guessing, until we see some evidence. Good thing I am good for at least a few more years (hopefully) before I have to face another update/upgrade decision.:D
 
Last edited:

enricoclaudio

Major Contributor
Audio Company
Joined
Jan 7, 2021
Messages
1,109
Likes
2,195
Location
Houston, TX - USA
The only thing I know is that over my dead body I would sell my X6700H purchased in January this year in exchange for the same model with replacement ESS DACs. Same applies to my Panasonic UB9000 with AKM DACs ;)
 

ShadowFiend

Member
Joined
Dec 20, 2020
Messages
69
Likes
89
@peng : Thanks for the info. I have not yet seen the new board of 4700 or 6700. My comment is only based on my experience on production of consumer products.

If they do not still use 8-channel 48 pin DAC, apart from low-end ESS like 9006, it can be Cirrus Logic CS4382A or CS4385A. Performance is suffered, of course, that's possibly a reason why Denon does not specify the new chip, to avoid uproar from customers, for example this forums.
 
OP
M

mike7877

Addicted to Fun and Learning
Joined
Aug 5, 2021
Messages
698
Likes
140
I've got the cover off, on the top board of the 3700 there is a PCM1803A. THD+n 93db @48khz
There are two PCM5100A's THD+n 90db beside it . The wires coming from the board right beside these three chips go to the amplifier boards, both the front and back boards. There are 13 wires.

The amplifier boards are also fed by something on the third board down from the top. My view is blocked so I can't give an accurate total or the names of the chips beside the pins, but I can say there are a few more wires going to the amps than there are coming from the top.
 

Jbrunwa

Senior Member
Joined
Mar 13, 2021
Messages
301
Likes
339
Location
Seattle
I've got the cover off, on the top board of the 3700 there is a PCM1803A. THD+n 93db @48khz
There are two PCM5100A's THD+n 90db beside it . The wires coming from the board right beside these three chips go to the amplifier boards, both the front and back boards. There are 13 wires.

The amplifier boards are also fed by something on the third board down from the top. My view is blocked so I can't give an accurate total or the names of the chips beside the pins, but I can say there are a few more wires going to the amps than there are coming from the top.
I just submitted a technical support request to Denon on my newly purchased AVR-X8500HA serial number to ask what DAC chips are installed. You might try asking them about yours.
 
OP
M

mike7877

Addicted to Fun and Learning
Joined
Aug 5, 2021
Messages
698
Likes
140
I just submitted a technical support request to Denon on my newly purchased AVR-X8500HA serial number to ask what DAC chips are installed. You might try asking them about yours.

Good idea. There are a lot of screws - one on each HDMI connector on the top board for starters. And cable ties in the way - I'm not sure I can get this apart without it being obvious I've been in it. I might need to return it..
 

peng

Master Contributor
Forum Donor
Joined
May 12, 2019
Messages
5,757
Likes
5,328
I've got the cover off, on the top board of the 3700 there is a PCM1803A. THD+n 93db @48khz
There are two PCM5100A's THD+n 90db beside it . The wires coming from the board right beside these three chips go to the amplifier boards, both the front and back boards. There are 13 wires.

The amplifier boards are also fed by something on the third board down from the top. My view is blocked so I can't give an accurate total or the names of the chips beside the pins, but I can say there are a few more wires going to the amps than there are coming from the top.

The two PCM5100A would be for Z2/Z3. Edited typo, meant to type 5100 not 5101
 
Last edited:
OP
M

mike7877

Addicted to Fun and Learning
Joined
Aug 5, 2021
Messages
698
Likes
140
The two PCM5101A would be for Z2/Z3.

The chips definitely had PCM5100, not 5101 printed on them

Edit: if I remember the datasheet correctly, 5100 is 90, 5101 is 92 5102 is 93. They're all similarly horrible. The noise is low from the snr spec, about -110? making almost all of the unwanted information in the signal harmonics. I don't see 32 tone test results for these chips anywhere, I bet they'd look horrific
 
Last edited:
Top Bottom