In your sand witch ELISA there are many steps to be taken care.You are getting the signal as compared to your negative control. That means no signal from the troponin. Each steps are  important. Your capture may not be coating properly because of coating buffer and/or its pH and concentration of protein. 2nd step Blocking, 1% BSA cold have blocked all the active site of your coating which could not give any signal, third concentration of detecting agent. Yes, incubation is also a factor for coating and detection. Through,  troponin is a complex protein you try to incubate the capture protein over night at 4 degree centigrade and other step like detection for 2 hrs.

please take care of the above steps hope you will get the result

What's the concentration of your coating and primary antibody?

And do you think that the amount of Troponin I in your sample is immense?

each of my capture and detection: 1 ug/mL, I tried the capture at 10 with 1 detection, but didn't get much different results, so I'm sticking to 1, and 1

I'm still working on the protein concentration, my range is 0.0001 to 0.1 ug/mL

Are you sure that your A's pair doesn't recognise the same epitope of protein? Abcam suggest the Mouse monoclonal 16A11 as pair for 19C7 Ab in Sandwich ELISA. Try to change the detection Ab,  using capture Ab in range 0.1-10ug/ml and detection -0.1-1ug/ml

1.     check Jelenas suggestion. For a two-site-binding assay you need to have on your analyte at least two different epitopes and two different antibodies recognizing them.

How to get some insights: check the data for this mabs, cell line, origin, …

Experimentally: bind troponin at 5 µg/mL (in 0.1 mol/L carbonate buffer) at your solid phase over night at 4 °C, (microtiter plate)

wash, now incubate with the unlabelled antibody in different concentrations, wash, incubate with the labelled. If you find a dependency on the concentration of the unlabelled, both recognizing the same epitope.

2.       Forget all this blocking voodoo. Coating as described at pH 9.6 in 0.1 mol/L carbonate buffer with 2 – 5 µg/mL overnight is mostly sufficient. The solid phase is overloaded anyhow (binding capacitiy is at 100 ng/cm²). Store the coated plates frozen until use. Blocking (if there are some free binding sites available) happened during the first washing with PBS Tween 20 by the detergent. Blocking is only necessary for commercial assays where storage of the test kit at room temperature for longer periods is required. If you will do really a blocking, use a protein, which is proven free of your antigen. Do not be surprized by checking BSA.

3.       It’s better to use an sample buffer which is closed to the samples you would like to analyse. In most cases PBS 0,1% v/v Tween 20 works fine as washing buffer. Add some digested Gelatine (1-3% v/v (available as plasma expander in a pharmacy) and some phenol red. So you will have constant protein conditions despite of the sample dilution. If you need to decrease the background you can increase the NaCl concentration by factor 2 (0,3 mol/L final concentration).

4.       For the first experiments, use the checkerboard titration with large steps. The normal detection limit of an assay depends only on the affinity of the solid phase antibody. Never on the incubation time, not that much on the incubation temperature (so use room temperature). The classical incubation times 60 – 90 min results from avoiding time and temperature gradients over the plate. This would results in higher (im)precisions.

Thanks for your ideas, the 2 antibodies recognise different epitopes, according to the company's manual.

I tried to skip the blocking step, but the background was high, so I probably will keep on blocking, I compared BSA vs superblock, they're both similar.

  I dilute the capture antibodies in PBS PH: 7.4, never tried the carbonate, would it make a difference?

*Please, Stephan, I don't understand what you mean by "use the checkerboard titration with large steps", could you explain that? Thanks

Checkerboard titration:

Row A - H: different antibody concentration for coating (start with the max of 10 µg/mL (and use the carbonate buffer, use normal plates instead of "high binding", binding happens via hydrophobic bindings to the polystyrene and this works better at pH 9 ... 9.8). Coating over night. No blocking. Washing and now

Column 1 - 12 for different sample concentrations. 1 = blank (incubation buffer only), 2 = 0,1 ng/mL , 3 = 0,4 ng/mL, 4  = 1,6 ng/mL, 5 = 6,4 ng/mL ... 12 = 104000 ng/mL (or whatever). High dilution factor 3 ... 5. 

Labelled antibody = constant = 1 µg/mL

Incubation times: 90 min

Incubation buffer: PBS, 0.1% Tween 20, 0.3 mol/L NaCl, 3% Gelafundin

Substrate: TMB

Stop: H2SO4

Evaluation: Search for the best P/N-Ratio (Positve at e.g. 2 µg/mL, negative = blank), with the prerequisite that the absorbance max is between 1.2 ... 2.0. Not higher, not lower.

You can use the same setup do optimize the buffer, the incubation times, ....

Only if this works well, you can add the blocking (but it's would not be required anymore. 

Another way to build up an assay would be the competitive version.

solid phase = //

washing = /

; = no washing

Solid phase antibody // sample / troponin-labelled (biotin, HRP, ...) / substrate; stop

or

Troponin // sample + labelled antibody / substrate; stop